UAV control method, medium and terminal based on air computing system optimization

Abstract

The present invention discloses a method, medium and terminal for controlling a drone based on optimization of an aerial computing system. The method comprises: obtaining grouping information of each ground node of the aerial computing system, and obtaining a preset first system optimization model for maximizing the minimum value of the computing rate of the aerial computing system; obtaining the optimal hovering position of the drone in each ground node group and the hovering time corresponding to each optimal hovering position according to the grouping information of the ground nodes and the first system optimization model, so as to determine the continuous hovering and flight trajectory of the drone; obtaining the grouping information of each ground node according to the continuous hovering and flight trajectory and the first system optimization model; continuing to perform the step of determining the continuous hovering and flight trajectory of the drone according to the grouping information of each ground node until the minimum value of the aerial computing rate meets the preset conditions. Based on this, the aerial computing rate is effectively improved on the basis of effectively avoiding the situation of "edge nodes".

Description

Technical Field

[0001] The present invention relates to the technical field of Internet of Things, and in particular to a drone control method, medium and terminal based on air computing system optimization. Background Art

[0002] With the rapid development of mobile communication technology in recent years, 5G communication can provide users with large bandwidth, massive connections and low-latency data exchange services, which has greatly promoted the evolution of IoT applications. Wireless data aggregation is a key link in IoT big data processing. It is a mathematical process that runs on a server and uses distributed data samples to perform functional calculations.

[0003] Aerial computing technology is widely regarded as a promising wireless data aggregation solution that can achieve data aggregation efficiently and quickly. Aerial computing technology combines data collection and computing functions, where the computing process can be completed in the "air". The aggregation center can directly receive the calculation results, improve data aggregation efficiency, and reduce latency. With the development of drone technology, drones have flexible mobility and controllability, and are used as aerial data aggregation centers in the field of aerial computing technology to achieve data collection and computing functions.

[0004] At present, the aerial computing system based on drones mainly focuses on whether there is information transmission distortion during the data collection process, but ignores the problem of whether the computing rate in the aerial computing network is limited. When there are ground nodes with poor channels in the aerial computing system built based on drones, the computing rate is extremely low or even zero, resulting in computing interruption. Therefore, even if the existing aerial computing network based on drones proposes corresponding optimization problems to minimize information transmission distortion, it cannot guarantee the normal operation of aerial computing.

[0005] Based on this, how to provide a UAV control method to solve the technical problem that the computing rate of the aerial computing system is reduced or zero due to the generation of "edge ground nodes" in the aerial computing system, thereby causing the aerial computing system to be unable to operate normally. Summary of the invention

[0006] The present invention provides a UAV control method, medium and terminal based on optimization of an aerial computing system, aiming to solve the technical problem in the prior art that the existence of "edge ground nodes" in the aerial computing system causes the computing rate of the aerial computing system to be reduced or zero, thereby causing the aerial computing system to be unable to operate normally.

[0007] The technical solution adopted by the present invention is as follows:

[0008] A method for controlling an unmanned aerial vehicle based on optimization of an aerial computing system, the method comprising:

[0009] Acquire grouping information of each ground node of the air computing system, and acquire a preset first system optimization model for maximizing the minimum value of the air computing rate;

[0010] The grouping information is used to indicate the grouping of the ground nodes in the ground node group; the first system optimization model is constructed by taking the maximization of the minimum value of the computing rate of the aerial computing system as the objective function, and taking the UAV flight trajectory constraint of the aerial computing system, the ground node grouping constraint, and the UAV flight time allocation constraint as the constraint conditions;

[0011] According to the grouping information of the ground nodes and the first system optimization model, an optimal hovering position of the UAV in each ground node group and a hovering time corresponding to each optimal hovering position are obtained;

[0012] and determining the continuous hovering and flight trajectory of the UAV according to the optimal hovering position of the UAV and the corresponding hovering duration;

[0013] Determining grouping information of each ground node according to the continuous hovering and flight trajectory and the first system optimization model;

[0014] Continue to perform the step of determining the continuous hovering and flight trajectory of the UAV according to the grouping information of each ground node and the first system optimization model until the minimum value of the aerial computing rate meets the preset conditions, and obtain the optimal flight trajectory of the UAV, the optimal ground node grouping, and the optimal UAV flight time allocation of the aerial computing system;

[0015] The flight of the drone is controlled according to the optimal flight trajectory of the drone, the optimal ground node grouping, and the optimal drone flight time allocation.

[0016] Optionally, the method further comprises:

[0017] The grouping information of the ground nodes is used as the grouping variable, the occupancy time of each ground node group is used as the time allocation variable, and the flight trajectory of the UAV is used as the flight trajectory variable;

[0018] Based on the grouping variable, the time allocation variable, the flight trajectory variable, the information transmission power of the ground node, and the additive white Gaussian noise power received by the UAV, an objective function for maximizing the minimum value of the air computing rate is constructed; and

[0019] Determine the flight trajectory constraints of the drone according to the initial point, the end point and the maximum flight speed of the drone flight;

[0020] Determining the ground node grouping attribution constraint according to the grouping information of the ground nodes;

[0021] Determining a UAV flight time allocation constraint based on the UAV flight time allocation;

[0022] The UAV flight trajectory constraints, ground node grouping constraints, and UAV flight trajectory constraints are combined according to the objective function to construct the first system optimization model for maximizing the minimum value of the aerial computing rate.

[0023] Optionally, obtaining the optimal hovering position of the UAV in each ground node group according to the grouping information of the ground nodes and the first system optimization model specifically includes:

[0024] Using a first auxiliary variable to transform the first system optimization model to obtain a second system optimization model;

[0025] Assuming that the mission duration of the UAV is greater than the corresponding preset threshold, the UAV flight trajectory constraints are ignored, and based on the grouping information of the ground nodes and the second system optimization model that ignores the UAV flight trajectory constraints, the optimal hovering position of the UAV in each ground node group and the hovering duration corresponding to each optimal hovering position are obtained.

[0026] Optionally, assuming that the mission duration of the drone is greater than a corresponding preset threshold, ignoring the drone flight trajectory constraint, and obtaining the optimal hovering position of the drone in each ground node group according to the grouping information of the ground nodes and the second system optimization model ignoring the drone flight trajectory constraint, specifically includes:

[0027] Assuming that the mission duration of the drone is greater than the corresponding preset threshold, the drone flight trajectory constraint in the second system optimization model is excluded from the constraint condition, and the grouping information of the ground nodes is used as a known item of the second system optimization model to obtain a third system optimization model;

[0028] Using the Lagrange dual method, determine the Lagrange dual variables;

[0029] Determining the Lagrangian function and the dual problem corresponding to the third system optimization model according to the Lagrangian dual variable;

[0030] The Lagrangian function is decomposed and the Lagrangian dual variables are iteratively optimized to obtain the optimal Lagrangian dual variables and the optimal hovering position of the UAV.

[0031] Optionally, obtaining grouping information of each ground node according to the continuous hovering and flight trajectory and the first system optimization model specifically includes:

[0032] Determining a drone flight trajectory and a drone flight time allocation for the drone based on the continuous hovering and flight trajectory;

[0033] Using the drone flight trajectory and the drone flight time distribution of the drone as known items of the second system optimization model to obtain a fourth system optimization model;

[0034] Using the second auxiliary variable and the third auxiliary variable, transforming the fourth system optimization model into a fifth system optimization model;

[0035] According to the continuous convex approximation method, the fifth system optimization model is processed to obtain the grouping information of each ground node.

[0036] Optionally, the fifth system optimization model is optimized according to the continuous convex approximation method to obtain grouping information of each ground node, specifically including:

[0037] The fifth system optimization model is processed according to the continuous convex approximation method to obtain a suboptimal solution of the fifth system optimization model;

[0038] The suboptimal solution is substituted into the fifth system optimization model, and the optimal solution of the fifth system optimization model is obtained through an integer programming solver, so as to determine the grouping information of each ground node according to the optimal solution of the fifth system optimization model.

[0039] Optionally, before obtaining the grouping information of each ground node of the aerial computing system, the method further includes:

[0040] Determine the number of ground node groups and the location information of each ground node;

[0041] generating a plurality of center points corresponding to the number of the groups in a ground area corresponding to the aerial computing system;

[0042] Calculate the distance between each ground node and each center point as the node-center distance;

[0043] According to the distance between each node and the center point, the ground node is assigned to the ground node group where the nearest center point is located;

[0044] According to the position information of the ground nodes of each ground node group, the new center point of the ground node group is determined, and the steps of calculating the distance between each ground node and each center point as the node-center point distance and allocating the ground node to the ground node group where the nearest center point is located according to each node-center point distance are performed until the center point of each ground node group no longer changes.

[0045] Optionally, before acquiring the node location information of each ground node in the aerial computing system, the method further includes:

[0046] Acquire a ground area corresponding to the aerial computing system and determine a midpoint of the ground area;

[0047] The midpoint of the ground area is used as the origin of the coordinate system, and the x-axis and y-axis of the coordinate system are both set on the ground;

[0048] Using the altitude of the drone as the z-axis of the coordinate system to construct the coordinate system of the aerial computing system;

[0049] According to the coordinate system, the position information of each ground node in the ground area is determined and stored.

[0050] A computer-readable storage medium stores one or more programs, and the one or more programs can be executed by one or more processors to implement the steps in drone control of an aerial computing system as described above.

[0051] A terminal comprises: a processor and a memory; the memory stores a computer-readable program that can be executed by the processor; when the processor executes the computer-readable program, the steps in any of the above-mentioned unmanned aerial vehicle control methods based on air computing system optimization are implemented.

[0052] Beneficial effects: Compared with the prior art, the present invention provides a method, medium and terminal for controlling a drone based on optimization of an aerial computing system. The method obtains the optimal hovering position of the drone in each ground node group and the hovering time corresponding to each optimal hovering position through the grouping information of each ground node of the aerial computing system and the first system optimization model, wherein the first system optimization model is used to maximize the minimum value of the aerial computing rate. Then, according to the optimal hovering position and the corresponding hovering time, the continuous hovering and flight trajectory of the drone is generated. Therefore, based on the generated continuous hovering and flight trajectory of the drone and the first system optimization model, the grouping information of each ground node is determined, that is, the grouping information of the ground node is root-updated to obtain new grouping information, and the continuous hovering and flight trajectory of the drone is continuously calculated according to the new grouping information, so as to iterate until the minimum value of the aerial computing rate meets the preset conditions, and the drone optimal flight trajectory, optimal ground node grouping belonging and optimal drone flight time allocation of the drone are used, and the drone flight is controlled according to the optimal flight trajectory of the drone, the optimal ground node grouping belonging and the optimal drone flight time allocation, so as to complete the data collection and calculation of the aerial computing system. That is to say, through the above technical solution, the optimal flight trajectory, optimal ground node grouping and optimal UAV flight time allocation can be obtained during the UAV flight mission, so that the computing rate of the ground node group with the worst computing rate is maximized, and the emergence of "edge ground nodes" is effectively avoided, thereby ensuring the normal operation of the aerial computing system. BRIEF DESCRIPTION OF THE DRAWINGS

[0053] Figure 1 A schematic diagram of the system structure of an aerial computing system based on a drone provided in this embodiment;

[0054] Figure 2 A schematic diagram of the structure of the drone provided in this embodiment;

[0055] Figure 3 A schematic diagram of the structure of the ground node provided in this embodiment;

[0056] Figure 4 A flowchart of a drone control method based on air computing system optimization provided in this embodiment;

[0057] Figure 5 A flowchart of a method for controlling a drone based on optimization of an aerial computing system provided in this embodiment;

[0058] Figure 6 A flowchart of a method for constructing a first system optimization model provided in this embodiment;

[0059] Figure 7A schematic diagram of a continuous hovering and flight trajectory and ground node deployment position provided in this embodiment;

[0060] Figure 8 A schematic diagram of performance comparison provided for this embodiment;

[0061] Fig. 9 A schematic diagram of the structure of a terminal provided in this embodiment. DETAILED DESCRIPTION

[0062] The present invention provides a drone control method, medium and terminal of an aerial computing system. In order to make the purpose, technical solution and effect of the present invention clearer and more specific, the present invention is further described in detail with reference to the accompanying drawings and examples. It should be understood that the specific embodiments described herein are only used to explain the present invention and are not used to limit the present invention.

[0063] It will be understood by those skilled in the art that, unless expressly stated, the singular forms "one", "said", and "the" used herein may also include plural forms. It should be further understood that the term "comprising" used in the specification of the present invention refers to the presence of the features, integers, steps, operations, elements and / or components, but does not exclude the presence or addition of one or more other features, integers, steps, operations, elements, components and / or groups thereof. It should be understood that when we refer to an element as being "connected" or "coupled" to another element, it may be directly connected or coupled to the other element, or there may be an intermediate element. In addition, the "connection" or "coupling" used herein may include wireless connection or wireless coupling. The term "and / or" used herein includes all or any unit and all combinations of one or more associated listed items.

[0064] It will be understood by those skilled in the art that, unless otherwise defined, all terms (including technical and scientific terms) used herein have the same meaning as generally understood by those skilled in the art in the art to which the present invention belongs. It should also be understood that terms such as those defined in general dictionaries should be understood to have meanings consistent with the meanings in the context of the prior art, and will not be interpreted with idealized or overly formal meanings unless specifically defined as herein.

[0065] We are now in the era of big data communication. In the future, nodes in the Internet of Things are expected to collect and use data from the environment, or communicate with each other to explore the meaning of data, such as temperature, pressure, health, and wireless channel status. Wireless data aggregation is a key link in the processing of big data in the Internet of Things. Wireless data aggregation is a mathematical process that runs on a server and uses distributed data samples to perform required functional calculations. Generally speaking, IoT nodes perform wireless data aggregation through a "communication first, then calculation" protocol, but in large-scale IoT applications (such as smart cities), especially when wireless network resources are scarce, data aggregation efficiency is very unsatisfactory, and it is even more challenging for delay-sensitive scenarios.

[0066] In order to overcome the above problems, air computing technology is widely regarded as a promising solution for wireless data aggregation, which can achieve data aggregation efficiently and quickly. In air computing, the aggregation center can use the signal superposition of multiple access channels to effectively balance the data transmitted from massive nodes in a simulated manner. Compared with the traditional "communication first, then calculation" protocol, air computing technology combines data collection and calculation functions, in which the calculation process can be completed in the "air", and the aggregation center can directly receive the calculation results.

[0067] The concept of air computing technology is to design structural coding to utilize signal interference in multiple access channels to achieve reliable function operations. In existing research, a simple uncoded signal transmission method has been proven to be able to achieve optimal function distortion minimization in the case of independent Gaussian sources. At present, there are methods to minimize the function distortion of air computing systems from the aspects of coding design, multi-antenna beamforming design, and power configuration. In other words, the current air computing technology focuses on signal transmission distortion, while the calculation rate of air computing systems is often ignored. The calculation rate of air computing depends on the node with the worst signal-to-noise ratio in the air computing system, and in extreme cases, the calculation rate may be zero, resulting in the inability of the air computing system to operate normally. In other words, in some cases, the calculation of the air computing system may be interrupted. Therefore, even if the corresponding optimization problem is proposed in the existing UAV-based air computing network to minimize information transmission distortion, the normal operation of the air computing system cannot be guaranteed.

[0068] In order to solve the above problems, in an embodiment of the present invention, the grouping information of each ground node of the aerial computing system is first obtained, and a preset first system optimization model for maximizing the minimum aerial computing rate is obtained; then, based on the grouping information of the ground nodes and the first system optimization model, the optimal hovering position of the UAV in each ground node group and the hovering time of the UAV at each optimal hovering position are obtained; thereby, according to the optimal hovering position and hovering time of the UAV, the continuous hovering and flight trajectory of the UAV are determined; finally, according to the continuous hovering and flight trajectory and the first system optimization model, the grouping information of each ground node of the UAV is determined; then, based on the obtained grouping information and the first system optimization model, the continuous hovering and flight trajectory of the UAV is determined until the minimum computing rate of the aerial computing system meets the preset conditions, thereby obtaining the optimal flight trajectory of the UAV, the optimal ground node grouping and the optimal UAV flight time allocation of the aerial computing system, and controlling the flight of the UAV according to the optimal flight trajectory of the UAV, the optimal ground node grouping and the optimal UAV flight time allocation information. It can be seen that through the technical solution provided by the embodiment of the present invention, the optimal flight trajectory of the UAV, the optimal ground node grouping and the optimal UAV flight time allocation can be obtained through cyclic iteration, so as to control the UAV to fly according to the optimal flight trajectory of the UAV, the optimal ground node grouping and the optimal UAV flight time allocation, avoid the occurrence of "ground edge nodes" causing the air computing rate of the air computing system to be reduced or zero, and improve the computing rate of the air computing system to ensure the normal operation of the air computing system.

[0069] The drone control method based on the optimization of the aerial computing system provided by the embodiment of the present invention is shown in FIG1. ​​The aerial computing system is composed of a drone (for example, a multi-rotor drone) and multiple ground IoT nodes (i.e., ground nodes). In the aerial computing system, the drone is used as a data aggregation center to collect and calculate data from multiple ground nodes through aerial computing aggregation. Since the controllable mobility of the drone can shorten the distance between the ground node and the aggregation center (i.e., the drone), a higher quality wireless channel environment is provided, which can effectively improve network performance.

[0070] like Figure 2 As shown in the figure, the UAV mainly includes: data input terminal, central control module, trajectory control module, communication control module, group selection module and wireless information receiving terminal. Figure 3 As shown, the ground node mainly includes: a data input terminal, a central control module, a communication control module, a group selection module and a wireless information sending terminal.

[0071] Among them, Figure 2As shown, the data input end of the drone receives the input location information of each ground node, and the central control module of the drone can group the ground nodes according to the location information, and obtain the optimization result of the drone flight planning after a series of optimization operations. Among them, the central control module of the drone also optimizes the group selection at each moment during the flight process, so that the drone can select the optimal group service mode at the current moment in real time when flying to different locations (that is, select one group of ground nodes for information aggregation). The optimization results output by the central control module of the drone can realize trajectory optimization and group selection through the trajectory control module, the communication control module and the group selection module. Finally, the wireless information receiving end of the drone can aggregate information.

[0072] like Figure 3 As shown, the optimal flight path planning of the UAV, the real-time location of the UAV and the node affiliation are input as input data to the central control module of the ground node through the data input terminal of the ground node. The central control module of the ground node can obtain the optimal selection result of whether to send information to the UAV after a series of optimization operations based on the current location of the UAV and the node grouping affiliation information. The ground node can select the optimal communication mode at the current moment according to the real-time location of the UAV during flight (i.e., choose to send wireless information to the UAV or remain in standby mode). The optimization result output by the central control module of the ground node is realized through the communication control module and the group selection module of the UAV to select the communication mode. Finally, the ground node sends information to the UAV through its wireless information sending terminal.

[0073] It should be noted that the algorithms run by the central control module of the drone and the central control module of the ground node are coordinated, that is, the real-time grouping selection of both parties is unified. At any time, when the drone selects one of the ground node groups for information aggregation, the ground nodes in the group are in the wireless information sending state, and the ground nodes of other groups remain in the standby state.

[0074] The invention is further described below by describing embodiments in conjunction with the accompanying drawings.

[0075] This embodiment provides a method for controlling a drone based on optimization of an aerial computing system, as shown in FIGS. 4 and 5 , and the method includes:

[0076] S401, obtaining grouping information of each ground node of the air computing system, and obtaining a preset first system optimization model for maximizing the minimum value of the computing rate of the air computing system.

[0077] The grouping information is used to indicate the affiliation of the ground node in the ground node group. The first system optimization model is constructed with the objective function of maximizing the minimum value of the computing rate of the aerial computing system and the constraints of the UAV flight trajectory of the aerial computing system, the affiliation of the ground node grouping, and the UAV flight time allocation information.

[0078] Specifically, before step S401, the location information of each ground node in the aerial computing system may be obtained first, and then the ground nodes may be grouped according to the node location information of each ground node to obtain several ground node groups and grouping information of each ground node, and store them.

[0079] Further, the number of groups of the ground node group can be determined first, and then a plurality of center points corresponding to the number of groups can be generated in the ground area corresponding to the aerial computing system; then the distance between each ground node and each center point is calculated as the node-center distance; then the ground node is assigned to the ground node group where the nearest center point is located according to the distance between each node-center point; then the new center point of the ground node group is determined according to the location information of the ground node of each ground node group; and the step of calculating the distance between each ground node and each center point as the node-center distance, and assigning the ground node to the ground node group where the nearest center point is located according to the distance between each node-center point, until the center point of each ground node group no longer changes, thereby obtaining a plurality of ground node groups and grouping variables of each ground node. Thus, each location node can be assigned to a suitable ground node group, effectively preventing the omission of certain ground nodes and causing the interruption of aerial computing. The number of groups of the above-mentioned ground node group can be randomly generated, or can be generated according to the size of the ground area corresponding to the aerial computing system, or can be generated according to user input, which is not specifically limited in the embodiments of the present application.

[0080] Furthermore, the ground area corresponding to the aerial computing system can be obtained first, and the midpoint of the ground area can be determined; then the midpoint of the ground area is used as the origin of the coordinate system, and the x-axis and y-axis of the coordinate system are set on the ground, and the height of the drone is used as the z-axis of the coordinate system to construct the coordinate system of the aerial computing system; then, based on the coordinate system, the position information of each ground node in the ground area is determined and stored.

[0081] It can be understood that the ground area corresponding to the above-mentioned aerial computing system can be obtained based on the position information of each ground node in the aerial computing system, and the ground area includes all ground nodes in the aerial computing system. For example, according to the distribution of ground nodes, a circle is drawn with one of the ground nodes as the center, and all ground nodes are included in the circle. The circle is the ground area corresponding to the aerial computing system. For another example, according to the distribution of ground nodes, a rectangle is drawn with the ground node as the center, and all ground nodes are included in the rectangle. The rectangle is the ground area corresponding to the aerial computing system. It should be noted that the ground area can be not only a regular shape, but also an irregular shape. It can be understood that when the ground area is a regular shape, it is more convenient to determine the center of the ground area. When the ground area is an irregular shape, the center of gravity of the ground area can be used as the center, which is not specifically limited in the embodiment of the present invention.

[0082] In addition, in the embodiment of the present invention, the location information of the ground node and the drone can uniformly use the coordinate system established above.

[0083] In the embodiment of the present invention, Figure 6 As shown, constructing a first system optimization model for maximizing the minimum value of the air computing rate can be achieved by the following steps:

[0084] S601, taking the grouping information of the ground nodes as the grouping variable, taking the occupancy time of each ground node group as the time allocation variable, and taking the flight trajectory of the UAV as the flight trajectory variable.

[0085] S602, constructing an objective function for maximizing the minimum value of the aerial computing rate based on the grouping variables, the time allocation variables, the flight trajectory variables, the information transmission power of the ground nodes, and the additive Gaussian white noise received by the UAV.

[0086] S603, determining the flight trajectory constraints of the UAV according to the initial point, the end point and the maximum flight speed of the UAV flight.

[0087] S604: Determine the ground node grouping constraints according to the ground node grouping information.

[0088] S605, determining the UAV flight time allocation constraint according to the UAV flight time allocation.

[0089] S606, constructing a first system optimization model for maximizing the minimum value of the aerial computing rate based on the UAV flight trajectory constraints, ground node grouping constraints, and UAV flight time allocation constraints according to the above objective function.

[0090] It should be noted that the execution order of the above steps S602-S605 can be adjusted at will and is not limited to the above execution order. For example, step S602 can be executed first, then step S605, and finally steps S603 and S604. This is not specifically limited in the embodiments of the present invention.

[0091] Specifically, Figure 1 The aerial computing system shown in the figure contains k∈K={1,…,K} ground nodes, each of which is deployed at a fixed position. According to the constructed three-dimensional coordinate system, the three-dimensional coordinates of the ground node k are defined as (x k ,y k ,0),w k =(x k ,y k ) is the horizontal coordinate of the ground node. k ,y k ,0) can be used as the location information of ground node k.

[0092] The duration of a UAV flight cycle is T, and the flight altitude is fixed at H. Assume that the flight cycle T∈(0,T] of the UAV is discretized into N time slots of equal duration δ=T / N. Since δ is small enough, the position of the UAV is assumed to be unchanged in each time slot. In the constructed coordinate system, the horizontal coordinate of the UAV is defined as q[n]=(x[n],y[n]), n∈N={1,…,N}. The maximum flight speed of the UAV is The drones and ground nodes in the aerial computing system can obtain each other's location information in real time and accurately.

[0093] The drone needs to start from a given starting point, the flight speed during the flight cannot exceed the maximum flight speed and reach the end point at the end of the mission. Therefore, the constraints of the drone's flight trajectory are:

[0094]

[0095] q[0]=q initial ,q[N]=q final ,

[0096] Among them, ||·|| is the Euclidean norm, q initial is the initial position of the UAV, q final is the final position of the drone,

[0097] Therefore, assuming that the mission duration T of the drone exceeds the preset threshold (that is, T is large enough), So that there is at least one flight path that satisfies the UAV flight trajectory constraints. At any time slot n∈N, the distance between the UAV and the ground node k is defined as:

[0098]

[0099] Considering that the wireless channel between the UAV and the ground node is dominated by the line-of-sight channel, the channel gain between the UAV and the ground node k is defined as:

[0100]

[0101] Among them, β0 is defined as the channel gain when the reference distance is 1 meter, and α≥2 is defined as the path loss factor when the reference distance is 1 meter.

[0102] In the embodiment of the present invention, K ground nodes may be grouped into M ground node groups, each of which has K ground node groups. m Ground nodes, Generally speaking, each ground node can only be assigned to one ground node group. In the embodiment of the present invention, a set of binary indicator variables is defined: To indicate the grouping of each ground node. k,m =1 means that ground node k belongs to the mth group; Δ k,m = 0 means that ground node k does not belong to the mth group. Therefore, we can get as well as

[0103] Based on this, the ground node grouping constraints can be obtained as follows:

[0104]

[0105]

[0106] Since the drone follows the time division multiple access method, it only serves the ground nodes in one ground node group at any time. Therefore, in any time slot n∈N, assuming that K in the ground node group m m The ground node occupancy time γ m [n]≥0 sends information to the drone, then

[0107] That is, the UAV flight time allocation constraint is:

[0108]

[0109] According to the calculation rate formula of aerial calculation, the average calculation rate formula of ground node group m during the entire flight cycle of the UAV is expressed as:

[0110]

[0111] Among them, H(f m ({Z i ,i∈K m}))=log2(L 2 K m ) is defined as the entropy value of the objective function of the ground node group m, f m (·) is the corresponding objective function, Z i is the transmission data of the ground node, L is the maximum value of the feasible sequence of the calculation results, ò is defined as an integer exceeding the preset threshold to eliminate the ground nodes that do not belong to group m, P is defined as the information transmission power of the ground node, σ 2 is the additive white Gaussian noise power received by the UAV information receiver.

[0112] In the case of this grouping strategy, the calculation of the global objective function of the entire air computing system can be obtained by processing the calculation results of all groups through background calculation, and this process can be ignored. It can be concluded that the calculation rate of the proposed air computing system depends on the minimum group calculation rate min m∈M R m ({q[n],γ m [n],Δ k,m}), which is the minimum value of the computing rate of the air computing system.

[0113] Based on this, under the constraints of UAV flight trajectory constraints, ground node grouping constraints and UAV flight time allocation constraints, by optimizing the UAV flight trajectory {q[n]}, ground node grouping constraints {Δ k,m}、UAV flight time allocation {γ m [n]}, maximizing the minimum value of the calculation rate of the air computing system m∈M R m ({q[n],γ m [n],Δ k,m}). Therefore, the first system optimization model constructed can be expressed as:

[0114] (P1):

[0115]

[0116] q[0]=q initial ,q[N]=q final ;

[0117]

[0118]

[0119]

[0120] S402, according to the grouping information of the ground nodes and the system optimization model, the optimal hovering position of the UAV in each ground node group and the hovering time corresponding to each optimal hovering position are obtained.

[0121] Specifically, the first system optimization model can be transformed by using the first auxiliary variable to obtain the second system optimization model; then, assuming that the mission duration of the UAV is greater than the corresponding preset threshold, ignoring the UAV flight trajectory constraints, and solving the second system optimization model based on the grouping information of the ground nodes and ignoring the UAV flight trajectory constraints to obtain the optimal hovering position of the UAV in each ground node group and the hovering duration corresponding to each optimal hovering position.

[0122] Furthermore, assuming that the mission duration of the UAV is greater than the corresponding preset threshold, the UAV flight trajectory constraint in the second system optimization model is excluded from the constraint conditions, and the grouping information of the ground nodes is used as the known item of the second system optimization model to obtain the third system optimization model; the Lagrangian dual method is used to determine the Lagrangian dual variables; according to the Lagrangian dual variables, the Lagrangian function and the dual problem corresponding to the third system optimization model are determined; the Lagrangian function is decomposed and the Lagrangian dual variables are iteratively optimized to obtain the optimal Lagrangian dual variables and the optimal hovering position of the UAV.

[0123] Furthermore, the system optimization model is transformed using the first auxiliary variable R, and the obtained second system optimization model is as follows:

[0124] (P1.1):

[0125]

[0126] Among them, assuming It can always be greater than 0. It can be seen that the first system optimization model and the second system optimization model are not convex optimization problems. The complex multivariate coupling and binary group indicator variables {Δ k,m} are difficult to handle, and traditional convex optimization solutions are no longer useful, making it difficult to obtain the optimal solution.

[0127] Since, given the ground node grouping affiliation, that is, obtaining the ground node grouping information, the grouping variable {Δ k,m}It has been determined that the flight trajectory of the drone and the time allocation of information transmission of each ground node group can be optimized at this time.

[0128] Assumptions is the node set of ground node group m, where Then the second system optimization model can be expressed as:

[0129] (P2):

[0130]

[0131]

[0132] q[0]=q initial ,q[N]=q final

[0133]

[0134] in, Define the horizontal coordinate of ground node k of ground node group m as w k,m ∈K m ,m∈M.

[0135] However, the above second system optimization model is still not a convex optimization problem and it is difficult to find the optimal solution.

[0136] Therefore, in the embodiment of the present invention, it can be considered that the mission duration T of the drone is long enough (that is, the mission duration of the drone is greater than the corresponding preset threshold, T→∞), and given the time slot duration, the number of time slots N→∞. In this case, the flight time of the drone between any positions is negligible, so the drone flight trajectory constraint can be ignored. Therefore, considering that the mission duration of the drone is long enough, the optimality of the second system optimization model depends on the hovering position and hovering duration of the drone, so the superposition on the time slot can be changed to the integral on time, so the third system optimization model can be expressed as:

[0137] (P2.1):

[0138]

[0139]

[0140]

[0141] Among them, constraints and indicate that the ground nodes in each ground node group should send information to the UAV at different times. That is, at any time t, γ m (t) = 1 means that the ground nodes in the ground node group m send information to the UAV; otherwise, the opposite is true.

[0142] Even though the third system optimization model (P2.1) is not a convex optimization problem, it satisfies the so-called “time-division condition”. Therefore, there is a strong duality between the third system optimization model (P2.1) and its corresponding dual problem, and the Lagrange dual method can be used to obtain the optimal solution of the third system optimization model (P2.1).

[0143] Specifically, define λ m ≥0,m∈M is the Lagrangian dual variable of the mth constraint in the constraint condition, defined as Then the partial Lagrangian of the third system optimization model (P2.1) is:

[0144]

[0145] The Lagrangian function of (P2.1) is:

[0146]

[0147] Theorem: In order for g(λ) to be bounded, then it must satisfy

[0148] Then, the dual problem of the third system optimization model (P2.1) can be expressed as:

[0149] (D2.1):min λ g(λ)

[0150]

[0151]

[0152] Since the strong duality holds, obtaining the optimal solution to the dual problem (D2.1) is equivalent to obtaining the optimal solution to the third system optimization model (P2.1). Therefore, first, given a feasible Lagrangian dual variable λ, the problem is solved to obtain g(λ), and then the Lagrangian dual variable is optimized to minimize g(λ).

[0153] The dual problem (D2.1) can be decomposed into the following sub-problems:

[0154]

[0155]

[0156]

[0157] in,

[0158]

[0159] It can be seen that the above problem consists of countless sub-problems, each of which corresponds to each instantaneous time t. Since the value of the optimal solution to the problem is always 0, the optimal solution R of the first auxiliary variable R * Can be any value.

[0160] Since the problem is the same at every instant in time t, t can be ignored for clarity. Thus, at every instant in time t, the variable {γ m} has M feasible solutions. By comparing the M feasible solutions, we can get the optimal {γ m *}.

[0161] When γ m =1,m∈M and , the problem can be expressed as:

[0162]

[0163] Using the non-decreasing property of symmetric functions, the optimal solution to the above problem can be obtained as follows:

[0164]

[0165] The optimal solution to the above problem can be obtained by two-dimensional exhaustive search method. (m) , and define its corresponding optimal value as φ m (q (m) ,λ m ). By comparing M optimal values, the following optimal solution can be obtained:

[0166] if Then the ground nodes in the ground node group m transmit information to the UAV, that is q * =q (m) .

[0167] It should be noted that if any two of the M optimal values ​​are equal, then the optimal solutions obtained are the optimal solutions to the problem.

[0168] Then, the ellipsoid method is used to obtain the optimal Lagrangian dual variable λ opt , and opt Substituting into the optimal solution obtained above, we can get Therefore, the optimal hovering position of M drones is obtained. Each ground node group corresponds to an optimal hovering position.

[0169] That is to say, given the grouping information of the ground nodes and considering that the mission duration of the drone is long enough, the drone flight trajectory constraints (including: maximum flight speed constraints, starting point constraints and end point constraints) can be ignored. Therefore, the maximization of the minimum value of the aerial calculation rate only depends on the hovering position of the drone and the corresponding hovering duration. When the drone flight trajectory constraints are ignored, the Lagrangian dual method can be used to obtain the optimal solution. Then, the Lagrangian dual variable is introduced, the Lagrangian function and the dual problem are established, and the optimal dual variable and the optimal hovering position of the drone are obtained by decomposing the Lagrangian function and iteratively optimizing the dual variable.

[0170] From the above, we can see that after obtaining the optimal hovering positions of M drones, we only need to allocate the hovering time of each optimal hovering position:

[0171]

[0172]

[0173]

[0174] Among them, τ m Defined as the optimal hovering position of the drone q (m) The optimal solution is It can be obtained by convex optimization algorithm.

[0175] S403: Determine the continuous hovering and flight trajectory of the UAV based on the optimal hovering position of the UAV and the corresponding hovering duration.

[0176] Specifically, the traveling salesman algorithm is used to determine the shortest flight path of the UAV according to the initial point position, the terminal position and the optimal hovering position of each ground node group of the UAV; then, the UAV flight time allocation is determined according to the hovering time corresponding to each optimal hovering position obtained above and the maximum flight speed of the UAV, so as to obtain the continuous hovering and flight trajectory of the UAV.

[0177] Furthermore, the UAV continuously hovers at the optimal hovering position along the shortest traversal sequence obtained by the traveling salesman algorithm, and reaches the next optimal hovering position at the maximum flight speed, and finally reaches the end point.

[0178] Among them, continuous hovering and flight trajectory include: the starting point position of the drone, the end point position, the optimal hovering position and the flight time between the optimal hovering positions.

[0179] S404, obtaining grouping information of each ground node according to the continuous hovering and flight trajectory and the first system optimization model.

[0180] Furthermore, based on the continuous hovering and flight trajectory, the UAV flight trajectory and the UAV flight time allocation of the UAV are determined; the UAV flight trajectory and the UAV flight time allocation of the UAV are used as known items of the second system optimization model to obtain the fourth system optimization model; the fourth system optimization model is transformed into a fifth system optimization model by using the second auxiliary variable and the third auxiliary variable; according to the continuous convex approximation method, the fifth system optimization model is processed to obtain the grouping information of each ground node.

[0181] Furthermore, according to the continuous convex approximation method, the fifth system optimization model is processed to obtain a suboptimal solution of the fifth system optimization model; the suboptimal solution is substituted into the fifth system optimization model, and the optimal solution of the fifth system optimization model is obtained through an integer programming solver, so as to determine the grouping information of each ground node according to the optimal solution of the fifth system optimization model.

[0182] Furthermore, the above continuous hovering and flight trajectory pairs can be used to obtain the UAV flight trajectory q[n] and the UAV flight time distribution {γ m [n]}, that is, given the UAV flight trajectory q[n] and the UAV flight time allocation {γ m [n]}, the fourth system optimization model can be expressed as:

[0183] (P3):

[0184]

[0185]

[0186]

[0187] Due to the complexity of the variables in the constraints and the binary grouping variables {Δ k,m}, given q[n] and {γ m The fourth system optimization model (P3) of [n]} is not a convex optimization problem.

[0188] Therefore, using the second auxiliary variable {A m} and the third auxiliary variable The fifth system optimization model (P3.1) can be expressed as:

[0189] (P3.1):

[0190]

[0191]

[0192]

[0193] The fifth system optimization model (P3.1) is still non-convex, but it can be approximated into a convex optimization problem using the continuous convex approximation method. The continuous convex approximation method is to approximate the non-convex problem into a convex optimization problem in an iterative manner to approximate a suboptimal solution. Assume that the variable after the j ≥ 0 iteration is and R (j) , using Taylor's first-order expansion formula, we have:

[0194]

[0195]

[0196] The above approximate result (ie, suboptimal solution) is introduced into the fifth system optimization model (P3.1), and the optimal solution of the fifth system optimization model can be obtained by an integer programming solver (eg, Mosek) to obtain the grouping information of each ground node.

[0197] S405, continue to execute the steps of determining the continuous hovering and flight trajectory of the UAV according to the grouping information of each ground node and the first system optimization model, until the minimum calculation rate of the aerial computing system meets the preset conditions, and obtain the optimal flight trajectory of the UAV, the optimal ground node grouping and the optimal UAV time allocation of the aerial computing system.

[0198] That is to say, after obtaining the new grouping information, continue to execute steps S401-S404, and iterate until the minimum value of the aerial computing rate meets the preset conditions, thereby obtaining the optimal flight trajectory of the drone, the optimal ground node grouping, and the optimal drone time allocation of the aerial computing system.

[0199] S406, controlling the flight of the UAV according to the optimal flight trajectory of the UAV, the optimal ground node grouping, and the optimal UAV flight time allocation.

[0200] By using the drone control method based on air computing system optimization provided by the embodiment of the present invention, the grouping and attribution of ground nodes and continuous hovering and flight trajectory are iteratively optimized, so that the minimum value of the computing rate of the air computing system gradually converges until the minimum value of the computing rate meets the preset conditions, that is, the minimum value of the air computing rate is maximized to obtain the optimal flight trajectory of the drone, the optimal ground node grouping and the optimal drone flight time allocation, so as to control the flight of the drone, effectively avoid the problem of "edge ground nodes" appearing in the air computing system, causing the computing rate to be reduced or zero, causing the air computing to be interrupted, effectively improve the computing rate of the air computing system, and ensure the normal operation of the air computing system.

[0201] For example, assume that an aerial computing system randomly deploys K = 40 ground nodes in a 2000 × 2000 m ground area, and the drone flies at an altitude of 50 m (i.e., the drone's flight altitude is 50 m). 2 =70dBm, α=2, P = 1W, L = 2, and the ground nodes are divided into M = 4 groups. The drone control method based on the optimization of the aerial computing system provided by the embodiment of the present invention is used to maximize the minimum value of the computing rate of the aerial computing system. Figure 7 As shown in Figure 1, the drone starts from the starting point, traverses the optimal hovering positions of the four ground node groups, and finally reaches the end point. Among them, the optimal hovering position q of each ground node group is (m) The drone control method provided by the embodiment of the present invention is obtained and will not be described in detail here. Figure 7 It can be seen that ground nodes can be assigned to a ground node group based on their own deployment locations, and nearby ground nodes can be effectively avoided from appearing as "edge nodes" and from being in an environment with poor channels, so as to ensure the performance of the aerial computing system.

[0202] like Figure 8 As shown in , when a direct flight trajectory is used for aerial calculation, that is, the UAV flies from the initial point to the midpoint at a constant speed, during the flight, the transmission time of each ground node group is optimized to maximize the minimum value of the calculation rate. Figure 8 It can be seen that the continuous hovering and flight trajectory and ground node grouping proposed in the embodiment of the present invention effectively improve the aerial computing rate by more than 50% compared with the direct flight trajectory, and as the mission duration T of the UAV increases, the performance upper limit of the UAV flight trajectory constraint can be ignored after all.

[0203] Based on the above-mentioned drone control method based on optimization based on an aerial computing system, the present invention also provides a computer-readable storage medium, which stores one or more programs, and the one or more programs can be executed by one or more processors to implement the steps in the drone control method based on optimization based on an aerial computing system described in the above-mentioned embodiment.

[0204] Based on the above-mentioned drone control method based on air computing system optimization, the present invention also provides a terminal, such as Fig. 9 As shown, it includes at least one processor (processor) 90; display screen 91; and memory (memory) 92, and may also include a communication interface (Communications Interface) 93 and a bus 94. Among them, the processor 90, the display screen 91, the memory 92 and the communication interface 93 can communicate with each other through the bus 94. The display screen 91 is configured to display a preset user guide interface in the initial setting mode. The communication interface 93 can transmit information. The processor 90 can call the logic instructions in the memory 92 to execute the method in the above embodiment.

[0205] In addition, the logic instructions in the memory 92 described above can be implemented in the form of software functional units and can be stored in a computer-readable storage medium when sold or used as an independent product.

[0206] The memory 92 is a computer-readable storage medium that can be configured to store software programs, computer executable programs, such as program instructions or modules corresponding to the methods in the embodiments of the present disclosure. The processor 90 executes the functional application and data processing by running the software programs, instructions or modules stored in the memory 92, that is, implementing the methods in the above embodiments.

[0207] The memory 92 may include a program storage area and a data storage area, wherein the program storage area may store an operating system and at least one application required for a function; the data storage area may store data created according to the use of the terminal, etc. In addition, the memory 92 may include a high-speed random access memory and may also include a non-volatile memory. For example, a variety of media that can store program codes, such as a USB flash drive, a mobile hard disk, a read-only memory (ROM), a random access memory (RAM), a disk or an optical disk, may also be a transient storage medium.

[0208] In addition, the specific process of loading and executing the plurality of instructions in the storage medium and the processor in the terminal has been described in detail in the method above, and will not be described here one by one.

[0209] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit it. Although the present invention has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the aforementioned embodiments, or make equivalent replacements for some of the technical features therein. However, these modifications or replacements do not deviate the essence of the corresponding technical solutions from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims

1. A drone control method based on air computing system optimization, characterized in that: The method comprises: Acquire grouping information of each ground node of the aerial computing system, and acquire a preset first system optimization model for maximizing the minimum value of the computing rate of the aerial computing system; The grouping information is used to indicate the grouping of the ground nodes in the ground node group; the first system optimization model is constructed by taking the maximization of the minimum value of the computing rate of the aerial computing system as the objective function, and taking the UAV flight trajectory constraint of the aerial computing system, the ground node grouping constraint, and the UAV flight time allocation constraint as the constraint conditions; According to the grouping information of the ground nodes and the first system optimization model, an optimal hovering position of the UAV in each ground node group and a hovering time corresponding to each optimal hovering position are obtained; and determining the continuous hovering and flight trajectory of the UAV according to the optimal hovering position of the UAV and the corresponding hovering duration; Determining grouping information of each ground node according to the continuous hovering and flight trajectory and the first system optimization model; The obtaining of grouping information of each ground node according to the continuous hovering and flight trajectory and the first system optimization model specifically includes: Determining a drone flight trajectory and a drone flight time allocation for the drone based on the continuous hovering and flight trajectory; Using the drone flight trajectory and the drone flight time distribution of the drone as known items of the second system optimization model to obtain a fourth system optimization model; Using the second auxiliary variable and the third auxiliary variable, transforming the fourth system optimization model into a fifth system optimization model; The second auxiliary variable is: ； The third auxiliary variable is: ； in, is the second auxiliary variable, Grouping and affiliation of ground nodes, is the total number of ground nodes, is the ground node, m is the ground node group, M is the total number of ground node groups, is the third auxiliary variable, is an integer that exceeds the preset threshold to eliminate groups that do not belong to the group The ground nodes, is the height between the UAV and the ground node, The flight trajectory of the drone. is the information transmission power of the ground node, is the additive white Gaussian noise power received by the drone information receiver, n is the time slot, N is the number of time slots; According to the continuous convex approximation method, the fifth system optimization model is processed to obtain grouping information of each ground node; Continue to perform the step of determining the continuous hovering and flight trajectory of the UAV according to the grouping information of each ground node and the first system optimization model until the minimum value of the aerial computing rate meets the preset conditions, and obtain the optimal flight trajectory of the UAV, the optimal ground node grouping, and the optimal UAV flight time allocation of the aerial computing system; The flight of the drone is controlled according to the optimal flight trajectory of the drone, the optimal ground node grouping, and the optimal drone flight time allocation.

2. The method for controlling a drone based on air computing system optimization according to claim 1, characterized in that: The method further comprises: The grouping information of the ground nodes is used as the grouping variable, the occupancy time of each ground node group is used as the time allocation variable, and the flight trajectory of the UAV is used as the flight trajectory variable; Based on the grouping variable, the time allocation variable, the flight trajectory variable, the information transmission power of the ground node, and the additive white Gaussian noise power received by the UAV, an objective function for maximizing the minimum value of the air computing rate is constructed; and Determine the flight trajectory constraints of the drone according to the initial point, the end point and the maximum flight speed of the drone flight; Determining the ground node grouping attribution constraint according to the grouping information of the ground nodes; Determining a UAV flight time allocation constraint based on the UAV flight time allocation; The UAV flight trajectory constraints, ground node grouping constraints, and UAV flight trajectory constraints are combined according to the objective function to construct the first system optimization model for maximizing the minimum value of the aerial computing rate.

3. The method for controlling a drone based on optimization of an airborne computing system according to claim 1, characterized in that: The obtaining, according to the grouping information of the ground nodes and the first system optimization model, the optimal hovering position of the UAV in each ground node group specifically includes: The first system optimization model is transformed by using the first auxiliary variable to obtain a second system optimization model, wherein the first auxiliary variable is ; Assuming that the mission duration of the UAV is greater than the corresponding preset threshold, the UAV flight trajectory constraints are ignored, and based on the grouping information of the ground nodes and the second system optimization model that ignores the UAV flight trajectory constraints, the optimal hovering position of the UAV in each ground node group and the hovering duration corresponding to each optimal hovering position are obtained.

4. The method for controlling a drone based on airborne computing system optimization according to claim 3, characterized in that: Assuming that the mission duration of the drone is greater than the corresponding preset threshold, ignoring the drone flight trajectory constraint, and according to the grouping information of the ground nodes and the second system optimization model ignoring the drone flight trajectory constraint, obtaining the optimal hovering position of the drone in each ground node group, specifically includes: Assuming that the mission duration of the drone is greater than the corresponding preset threshold, the drone flight trajectory constraint in the second system optimization model is excluded from the constraint condition, and the grouping information of the ground nodes is used as a known item of the second system optimization model to obtain a third system optimization model; Using the Lagrange dual method, determine the Lagrange dual variables; Determining the Lagrangian function and the dual problem corresponding to the third system optimization model according to the Lagrangian dual variable; The Lagrangian function is decomposed and the Lagrangian dual variables are iteratively optimized to obtain the optimal Lagrangian dual variables and the optimal hovering position of the UAV.

5. The method for controlling a drone based on air computing system optimization according to claim 1, characterized in that: The fifth system optimization model is processed according to the continuous convex approximation method to obtain the grouping information of each ground node, specifically including: The fifth system optimization model is processed according to the continuous convex approximation method to obtain a suboptimal solution of the fifth system optimization model; The suboptimal solution is substituted into the fifth system optimization model, and the optimal solution of the fifth system optimization model is obtained through an integer programming solver, so as to determine the grouping information of each ground node according to the optimal solution of the fifth system optimization model.

6. The method for controlling a drone based on optimization of an airborne computing system according to claim 1, characterized in that: Before obtaining the grouping information of each ground node of the aerial computing system, the method further includes: Determine the number of ground node groups and the location information of each ground node; generating a plurality of center points corresponding to the number of the groups in a ground area corresponding to the aerial computing system; Calculate the distance between each ground node and each center point as the node-center distance; According to the distance between each node and the center point, the ground node is assigned to the ground node group where the nearest center point is located; According to the position information of the ground nodes of each ground node group, the new center point of the ground node group is determined, and the steps of calculating the distance between each ground node and each center point as the node-center point distance and allocating the ground node to the ground node group where the nearest center point is located according to each node-center point distance are performed until the center point of each ground node group no longer changes.

7. The method for controlling a drone based on optimization of an airborne computing system according to claim 1, characterized in that: Before acquiring the node location information of each ground node in the aerial computing system, the method further includes: Acquire a ground area corresponding to the aerial computing system and determine a midpoint of the ground area; The midpoint of the ground area is used as the origin of the coordinate system, and the x-axis and y-axis of the coordinate system are both set on the ground; Using the altitude of the drone as the z-axis of the coordinate system to construct the coordinate system of the aerial computing system; According to the coordinate system, the position information of each ground node in the ground area is determined and stored.

8. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores one or more programs, and the one or more programs can be executed by one or more processors to implement the steps in the drone control method based on air computing system optimization as described in any one of claims 1 to 7.

9. A terminal, characterized in that: include: A processor and a memory; the memory stores a computer-readable program that can be executed by the processor; when the processor executes the computer-readable program, the steps in the drone control method based on air computing system optimization as described in any one of claims 1 to 7 are implemented.

Images