A method and system for unmanned aerial vehicle cooperative edge computing offloading applied to a power terminal

Abstract

The application discloses a kind of unmanned aerial vehicle cooperation edge computing unloading method and system applied to electric power terminal, method includes: obtaining each electric power equipment parameter, unmanned aerial vehicle parameter and multi-access edge computing parameter;Calculate the time delay generated by multi-access edge computing system and each link of electric power equipment except task queuing delay;The order of electric power task in waiting queue is adjusted using priority dominated queue real-time scheduling algorithm, and task queuing delay is obtained;Calculate the total time of processing data between multi-access edge computing system and electric power equipment;Unmanned aerial vehicle cooperation edge computing unloading model applied to electric power terminal is constructed;The model is solved using differential evolution algorithm of hybrid simulated annealing algorithm, and unmanned aerial vehicle cooperation edge computing unloading scheme applied to electric power terminal is obtained.The application provides unmanned aerial vehicle to assist multi-access edge computing, so that unmanned aerial vehicle can cover complex environment, meet the supporting role to electric power equipment.

Description

TECHNICAL FIELD

[0001] The application relates to a UAV cooperative edge computing offloading method and system applied to a power terminal and belongs to the technical field of power system communication. BACKGROUND

[0002] With the update iteration of mobile network technology, powerful communication capabilities support the arrival of the era of interconnection and intercommunication of all things. The power system introduces the Internet of Things to create and realize scenes such as power parks and intelligent substations. However, as the scale of the power system becomes increasingly complex and large, its requirements for the Internet of Things do not stop at interconnection and intercommunication. Due to the needs of social development and industrial growth, power scenarios need to evolve in the direction of "intelligentization". Power equipment needs to have or be supported by other equipment with data processing capabilities to respond to any situation in real time. Such a process cannot achieve real-time and intelligent performance and cannot meet the basic requirements of low latency and high reliability. Therefore, multi-access edge computing is introduced into the power system. Before the popularization of multi-access edge computing, the data center described above is generally served by a cloud computing server.

[0003] Cloud computing is known for its powerful computing capabilities, but its volume is large and far away from the machine. Although it can accurately process massive data, power equipment has lost the ability to quickly respond to situations. Multi-access edge computing is essentially deploying a small cloud computing server near the equipment to compress transmission time by reducing physical distance, so that the data collected by the equipment can be processed simultaneously by itself and the small cloud server, thereby achieving fast data processing and making a response.

[0004] Generally, multi-access edge computing servers are deployed near base stations, and the coverage of their capabilities is fixed. However, when a fixed-position multi-access edge computing server faces situations such as remote location, complex terrain environment, and widely distributed and sparse equipment, it is helpless. From the perspectives of communication capability and economic cost, a means is needed to deploy multi-access edge computing services in places that cannot be reached by traditional solutions or have high costs.

[0005] Deploying multi-access edge computing on a UAV is the best solution to this problem. The UAV has the advantages of flexible deployment, rapid maneuvering, data transmission through antennas, good communication conditions, high cost performance, and server carrying, so it can cover multi-access edge computing services to complex terrain environments.

[0006] Therefore, a UAV cooperative edge computing offloading method applied to a power terminal is the direction of current research. SUMMARY

[0007] The unmanned aerial vehicle cooperative edge computing offloading method applied to the power terminal can provide unmanned aerial vehicle assisted multi-access edge computing, enable the unmanned aerial vehicle to cover complex environments, and meet the support for power equipment.

[0008] In a first aspect, the present application provides an unmanned aerial vehicle cooperative edge computing offloading method applied to a power terminal, comprising:

[0009] Obtaining each power equipment parameter, unmanned aerial vehicle parameter and multi-access edge computing parameter;

[0010] Based on the obtained parameters, the time delay generated between the multi-access edge computing system and the power equipment in each link except the task queuing time delay is calculated, including the transmission time delay d_sen, the multi-access edge computing processing time delay d_cal, the power terminal processing time delay d_equ and the unmanned aerial vehicle relay time delay d_re; the priority dominant queue real-time scheduling algorithm is used to adjust the order of the power task in the waiting queue, and the task queuing time delay d_q is obtained;

[0011] Based on the time delay, the total time for processing data between the multi-access edge computing system and the power equipment is calculated;

[0012] Based on the shortest total time, an unmanned aerial vehicle cooperative edge computing offloading model applied to the power terminal is constructed;

[0013] The differential evolution algorithm of the hybrid simulated annealing algorithm is used to solve the unmanned aerial vehicle cooperative edge computing offloading model applied to the power terminal, and an unmanned aerial vehicle cooperative edge computing offloading scheme applied to the power terminal is obtained.

[0014] In combination with the first aspect, further, the time delay generated between the multi-access edge computing system and the power equipment in each link except the task queuing time delay comprises:

[0015] The multi-access edge computing processing time delay d_cal is calculated by the following formula:

[0016]

[0017] In formula (1), g is the multi-access edge computing offloading ratio; D is the data collected by the power terminal; E_cal is the computing power of the unmanned aerial vehicle itself;

[0018] The power terminal processing time delay d_equ is calculated by the following formula:

[0019]

[0020] In formula (2), E_equ is the computing power of the power terminal itself;

[0021] The transmission delay d_sen is calculated by the following formula:

[0022]

[0023] In formula (3), B is the bandwidth used by the power terminal; P is the power; q is the antenna gain; w is the channel gain, w0 is the unit distance channel gain; l EU is the distance between the power terminal and the unmanned aerial vehicle that offloads data for it; σ 2 is the noise power;

[0024] In the multi-access edge computing system formed by the unmanned aerial vehicle, the flight rule is a relative position fixed, distance equal, and the unmanned aerial vehicle flight trajectory is a clockwise circle. The relay delay d_re of the unmanned aerial vehicle is calculated by the following formula:

[0025]

[0026] In formula (4), R is the flight radius of the unmanned aerial vehicle; A is the number of unmanned aerial vehicles; a is the position sequence number of the unmanned aerial vehicle closest to the power device in the clockwise direction relative to the unmanned aerial vehicle that transmits the calculation result; v is the flight speed of the unmanned aerial vehicle; The maximum delay d_wait that the power terminal needs to wait for the unmanned aerial vehicle to communicate with it is calculated by the following formula:

[0027]

[0028] In formula (5), r is the coverage radius of the communication capability of the unmanned aerial vehicle.

[0029] In combination with the first aspect, further, the priority-based queue real-time scheduling algorithm is used to adjust the order of the power task in the waiting queue, which includes:

[0030] It is judged whether there is a task in the queue. If there is no task, the current task is directly added to the queue. If there is a task, the last task in the queue is replaced in order;

[0031] It is judged whether the current task will cause the delay of the task behind it after being inserted. If it will cause the delay, the task of the same level that can be replaced is found from the insertion position, and if it cannot be found, the current task is stopped from being inserted, and the algorithm is stopped;

[0032] If the current task does not reach the same level task or after the higher level task, but stops moving forward because it meets the delay requirement, the replacement cycle is ended. If it does not meet the delay requirement, but the current task reaches the same level task and stops moving forward, the replacement cycle is ended.

[0033] In combination with the first aspect, further, the total time for processing data between the multi-access edge computing system and the power device is calculated by the following formula:

[0034] T_tal = max{d_equ, d_MEC}

[0035] = max{d_equ, d_sen + d_q + d_cal + d_qr + d_re + d_wait + d_d} (6)

[0036] In formula (6), T_tal is the total time for the multi-access edge computing system and the power terminal to process data; d_equ is the power terminal processing delay; d_sen is the transmission delay; d_q is the task queuing delay; d_cal is the multi-access edge computing processing delay; d_qr is the queuing relay delay; d_re is the unmanned aerial vehicle relay delay; d_wait is the maximum delay that the power terminal needs to wait until the unmanned aerial vehicle can communicate with it; and d_d is the delay of the data transmission.

[0037] In combination with the first aspect, further, the unmanned aerial vehicle cooperative edge computing offloading model applied to the power terminal is constructed by the following formula:

[0038]

[0039] c.t.t n ≤ E n

[0040] 0 ≤ b n ≤ 1

[0041] 0 ≤ β n ≤ 1

[0042]

[0043]

[0044]

[0045]

[0046]

[0047]

[0048] In formula (7), n is the power terminal serial number, N is the number of power terminals; t n is the total time for the multi-access edge computing system and the power terminal n to process data, E n is the effective time for the multi-access edge computing system and the power terminal n to process data; b n is the channel proportion occupied by the power terminal n; β n is the proportion of computing resources used by the power terminal n; and xa，n is the difference between the horizontal coordinates of the power terminal n and the UAV relative coordinates; y a，n is the difference between the vertical coordinates of the power terminal n and the UAV relative coordinates; g n is the offloading ratio of the power terminal n; D n is the data collected by the power terminal n; B0 is the downlink channel bandwidth; B1 is the uplink channel bandwidth; P n is the power of the power terminal n; q n is the antenna gain of the power terminal n; l n is the distance between the power terminal n and the UAV that offloads data; θ a is the angle between the coverage range of the UAV closest to the power terminal and the power terminal, when the power terminal is in the coverage range, θ a = 0.

[0049] In combination with the first aspect, further, the optimization step utilizes the differential evolution algorithm of the hybrid simulated annealing algorithm to solve the UAV cooperative edge computing offloading model applied to the power terminal, comprising:

[0050] Randomly generate X 3N-dimensional vectors in the feasible region Z to form the first generation population;

[0051] Through mutation, according to the scaling factor F, calculate the mutant individual m i of the ith individual x i in the population;

[0052] Through crossover, using the ith individual and the mutant individual, according to the crossover probability cr, calculate the trial individual n i ;

[0053] Through selection, determine the new child individual;

[0054] Form a new individual h i by exchanging any two parts of the existing individual;

[0055] Calculate the fitness increment Δ i ;

[0056] Calculate the selection probability l i ;

[0057] Determine the value of the individual x i through the calculation probability of the fitness increment;

[0058] Repeat the above steps until the maximum number of iterations is reached.

[0059] The second aspect of the present application provides a UAV cooperative edge computing offloading system applied to a power terminal, comprising:

[0060] An acquisition module is configured to acquire each power equipment parameter, UAV parameter and multi-access edge computing parameter;

[0061] A time delay calculation module is configured to calculate the time delay generated in each link between the multi-access edge computing system and the power equipment, including transmission time delay d_sen, multi-access edge computing processing time delay d_cal, power terminal processing time delay d_equ and UAV relay time delay d_re, based on the acquired parameters, and to adjust the order of the power task in the waiting queue by using a priority-oriented queue real-time scheduling algorithm, and obtain the task queuing time delay d_q.

[0062] A first calculation module is configured to calculate the total time for processing data between the multi-access edge computing system and the power equipment based on the time delay.

[0063] A model construction module is configured to construct a UAV cooperative edge computing offloading model applied to the power terminal based on the shortest total time.

[0064] A second calculation module is configured to solve the UAV cooperative edge computing offloading model applied to the power terminal by using a differential evolution algorithm of a hybrid simulated annealing algorithm, and obtain a UAV cooperative edge computing offloading scheme applied to the power terminal.

[0065] In a third aspect, the present application provides a computing device, characterized by comprising a processor and a storage medium.

[0066] The storage medium is configured to store instructions.

[0067] The processor is configured to operate according to the instructions to perform the steps of the method of the first aspect.

[0068] In a fourth aspect, the present application provides a computer-readable storage medium having a computer program stored thereon, characterized in that the program is executed by a processor to implement the steps of the method of the first aspect.

[0069] Compared with the prior art, the UAV cooperative edge computing offloading method and system applied to the power terminal provided by the embodiments of the present application have the following beneficial effects:

[0070] The present application acquires each power equipment parameter, UAV parameter and multi-access edge computing parameter, and uses a fixed-wing UAV to carry an edge computing server as an edge node for dynamic deployment, which can cover complex environments and greatly improve the application range of the multi-access edge computing system.

[0071] The application calculates the time delay generated in each link between the multi-access edge computing system and the power equipment, including transmission time delay d_sen, multi-access edge computing processing time delay d_cal, power terminal processing time delay d_equ and unmanned aerial vehicle relay time delay d_re, based on the obtained parameters, in addition to the task queuing time delay; the queuing of the power task in the waiting queue is adjusted by using the priority dominant queue real-time scheduling algorithm, and the task queuing time delay d_q is obtained; the queuing of the power task in the waiting queue is flexibly adjusted by using the priority dominant queue real-time scheduling algorithm, so that the high-priority task can be processed faster while meeting the time delay requirements of all power equipment;

[0072] The application calculates the total time for processing data between the multi-access edge computing system and the power equipment based on the time delay; and constructs an unmanned aerial vehicle cooperative edge computing offloading model applied to the power terminal based on the shortest total time;

[0073] The differential evolution algorithm of the hybrid simulated annealing algorithm is used to solve the unmanned aerial vehicle cooperative edge computing offloading model applied to the power terminal, so as to obtain an unmanned aerial vehicle cooperative edge computing offloading scheme applied to the power terminal. The differential evolution algorithm is combined with the simulated annealing algorithm and priority adjustment, so that better performance is obtained, and the purpose of minimizing the total time delay and processing the high-priority task first is achieved under the condition of meeting the time delay requirements of all power equipment. BRIEF DESCRIPTION OF DRAWINGS

[0074] Figure 1 It is a scene diagram of the unmanned aerial vehicle cooperative edge computing offloading method applied to the power terminal provided by the embodiment one of the application. DETAILED DESCRIPTION

[0075] The application will be further described below with reference to the drawings. The following embodiments are only used to more clearly illustrate the technical solutions of the application, and cannot be used to limit the protection scope of the application.

[0076] Embodiment one:

[0077] The flexibility and high load of the fixed-wing unmanned aerial vehicle are used to support the use of the power equipment in the environment with wide range, sparse power equipment distribution and difficult communication, and a kind of unmanned aerial vehicle cooperative multi-access edge computing offloading algorithm for different priority power equipment is proposed by combining real-time scheduling algorithm and differential evolution algorithm, and an unmanned aerial vehicle cooperative edge computing offloading method applied to the power terminal is constructed.

[0078] The system model used by the application is as follows: Figure 1As shown, the scenario model is composed of a multi-access edge computing server carried by a UAV and power equipment. The multi-access edge computing server carried by the UAV has data transmission and computing capability, and can process power service data. Meanwhile, the flexible mobility of the UAV can provide support for power equipment in an environment with a wide range, sparse power equipment distribution and difficult communication.

[0079] In terms of UAV selection, UAVs are divided into fixed-wing and rotary-wing types. Currently, the commonly seen rotary-wing UAVs are convenient to control and can hover, but have short endurance time, poor mobility and stability, low load capacity and low efficiency. The fixed-wing UAVs do not have the above-mentioned shortcomings, but do not have the hovering function, so they cannot stay in a fixed place to continuously provide communication and computing capability for the equipment. Because of the mobility of the fixed-wing UAVs, multiple UAVs are needed to provide stable and timely communication and computing capability, so as to avoid long waiting delay caused by a single fixed-wing UAV. The business demands of the power grid are various and different in importance, which is reflected in different requirements for time delay. Therefore, the importance of the tasks uploaded by the power equipment is distinguished by serial numbers according to the set business priority, and the higher the priority, the smaller the serial number, so as to determine the processing order when multiple task processing conflicts. The power business is divided into three priorities {p0, p1, p2} in the embodiment of the application to correspond to the three main situations.

[0080] The power equipment will experience the steps of uploading data to the UAV, queuing for processing, computing processing, queuing for relay, relay, waiting for the UAV to move, and downloading from the UAV, from uploading data to the UAV to obtaining the result. Because the computing result is generally small, the multi-access edge computing system of the present application ignores the time delay d_d of the UAV to the power equipment for downloading data and the queuing relay time delay d_r encountered by the UAV processing data to the nearest UAV to the power equipment.

[0081] The embodiment of the application provides a UAV cooperative edge computing offloading method applied to a power terminal, which comprises the following steps:

[0082] Obtaining each power equipment parameter, UAV parameter and multi-access edge computing parameter;

[0083] Based on the obtained parameters, the time delay generated between each link of the multi-access edge computing system and the power equipment except for the task queuing time delay is calculated, including the transmission time delay d_sen, the multi-access edge computing processing time delay d_cal, the power terminal processing time delay d_equ and the UAV relay time delay d_re. The priority-based queue real-time scheduling algorithm is used to adjust the order of the power task in the waiting queue, and the task queuing time delay d_q is obtained.

[0084] Based on the delay, the total time for processing data between the multi-access edge computing system and the power equipment is calculated;

[0085] Based on the shortest total time, a UAV cooperative edge computing offloading model applied to the power terminal is constructed;

[0086] The differential evolution algorithm with hybrid simulated annealing algorithm is used to solve the UAV cooperative edge computing offloading model applied to the power terminal, and a UAV cooperative edge computing offloading scheme applied to the power terminal is obtained.

[0087] The specific steps include:

[0088] Step 1: Obtain each power equipment parameter, UAV parameter and multi-access edge computing parameter.

[0089] The multi-access edge computing offloading ratio is set to g, the data collected by the power equipment is D, and the self-computing power of the power equipment is E_equ. Because of the mobility of the fixed-wing UAV, the power equipment cannot always be within the coverage range of the UAV communication capability, so the amount of data uploaded by the power equipment is limited. Assuming that the flight radius of the UAV is R, the coverage radius is r, and the flight speed is v, the upper limit of the data transmission from the power equipment to the UAV is:

[0090]

[0091] In formula (1), V is the channel capacity between the power equipment and the UAV.

[0092] Step 2: Based on the parameters obtained in step 1, the delay generated in each link between the multi-access edge computing system and the power equipment except the task queuing delay is calculated.

[0093] The multi-access edge computing server carried by the UAV calculates the multi-access edge computing processing delay d_cal through the following formula:

[0094]

[0095] In formula (2), g is the multi-access edge computing offloading ratio; D is the data collected by the power terminal; E_cal is the self-computing power of the UAV.

[0096] The power terminal processing delay d_equ is calculated through the following formula:

[0097]

[0098] In formula (3), E_equ is the self-computing power of the power terminal.

[0099] The transmission delay d_sen is calculated through the following formula:

[0100]

[0101] In formula (4), B is the bandwidth used by the power terminal device; P is the power; q is the antenna gain; w is the channel gain, w0 is the unit distance channel gain; l EU is the distance between the power terminal device and the unmanned aerial vehicle which offloads data for it; σ 2 is the noise power.

[0102] In the multi-access edge computing system constituted by the unmanned aerial vehicle, the flight rule is the relative position is fixed and the distance is equal, and the unmanned aerial vehicle flight trajectory is a clockwise circle. The d_re relay time in the unmanned aerial vehicle relay is calculated by the following formula:

[0103]

[0104] In formula (5), R is the unmanned aerial vehicle flight radius; A is the number of unmanned aerial vehicles; a is the position sequence number of the unmanned aerial vehicle closest to the power device in the clockwise direction relative to the unmanned aerial vehicle transmitting the calculation result; v is the unmanned aerial vehicle flight speed; the maximum delay d_wait that the power terminal device needs to wait until the unmanned aerial vehicle can communicate with it is calculated by the following formula:

[0105]

[0106] In formula (6), r is the coverage radius of the unmanned aerial vehicle communication capability.

[0107] Step 3: Adjust the order of the power task in the waiting queue by using the priority dominant queue real-time scheduling algorithm, and obtain the task queuing delay d_q.

[0108] In order to meet the requirement that the task with high priority is completed first, the multi-access edge computing system needs and can only adjust the waiting processing task queue after the power device offloads data to the unmanned aerial vehicle according to the priority, delay constraint and other factors. It calculates the last position of the task in the queue at this time which just meets the delay requirement of the task according to the time constraint of the task as an upper limit, the priority of the task, the current queue length and the elements as initial conditions, and inserts it. Then continuously exchange positions with low priority tasks until it meets the task with the same priority and stops. The priority dominant queue real-time scheduling algorithm adjusts the order of the power task in the waiting queue flexibly, meets the delay requirement of all power devices, and makes the high priority task be processed faster.

[0109] The priority dominant queue real-time scheduling algorithm is as follows:

[0110]

[0111]

[0112]

[0113] The steps are as follows:

[0114] Step 3.1: Determine whether there is a task in the queue. If there is no task, directly add the current task to the queue; if there is a task, sort and replace from the last position in the queue;

[0115] Step 3.2: Determine whether the current task will cause delay of the task behind it after insertion; if it will cause delay, find a task of the same level that can be replaced before the insertion position and replace it; if not found, stop inserting the current task and stop the algorithm;

[0116] Step 3.3: If the current task has not reached the same level task or after the higher level task, but stops moving forward because it meets the time delay requirement, end the replacement cycle; if it does not meet the time delay requirement, but the current task reaches the same level task and stops moving forward, end the replacement cycle.

[0117] Step 4: Based on the time delay, calculate the total time for processing data between the multi-access edge computing system and the power equipment.

[0118] The total time is calculated by the following formula:

[0119] T_tal=max{d_equ,d_MEC}

[0120] =max{d_equ,d_sen+d_q+d_cal+d_qr+d_re+d_wait+d_d} (7)

[0121] In formula (7), T_tal is the total time for processing data between the multi-access edge computing system and the power equipment; d_equ is the power terminal processing time delay; d_sen is the transmission time delay; d_q is the task queuing time delay; d_cal is the multi-access edge computing processing time delay; d_qr is the queuing relay time delay; d_re is the unmanned aerial vehicle relay time delay; d_wait is the maximum time delay that the power terminal needs to wait until the unmanned aerial vehicle can communicate with it; d_d is the time delay of the downlink data.

[0122] Step 5: Based on the shortest total time, build an unmanned aerial vehicle cooperative edge computing offloading model applied to the power terminal.

[0123] Under the condition of meeting all task time delay requirements, the multi-access edge computing system completes all tasks with the least time, while the high-priority power task is processed as early as possible, and the unmanned aerial vehicle cooperative edge computing offloading model applied to the power terminal is built, which is represented by the following formula:

[0124]

[0125] c.t.tn ≤E n

[0126] 0≤b n ≤1

[0127] 0≤β n ≤1

[0128]

[0129]

[0130]

[0131]

[0132]

[0133]

[0134] In formula (8), n is the serial number of the power terminal, N is the number of power terminals; t n is the total time for processing data between the multi-access edge computing system and the power terminal n, E n is the effective time for processing data between the multi-access edge computing system and the power terminal n; b n is the proportion of the channel occupied by the power terminal n; β n is the proportion of the computing resource used by the power terminal n; x a，n is the difference of the horizontal coordinate in the relative coordinates between the power terminal n and the unmanned aerial vehicle, y a，n is the difference of the vertical coordinate in the relative coordinates between the power terminal n and the unmanned aerial vehicle; g n is the offloading proportion of the power terminal n; D n is the data collected by the power terminal n; B0 is the downlink channel bandwidth; B1 is the uplink channel bandwidth; P n is the power of the power terminal n; q n is the antenna gain of the power terminal n; l n is the distance between the power terminal n and the unmanned aerial vehicle to which the data is offloaded; θ a is the angle between the coverage range of the unmanned aerial vehicle closest to the power terminal and the power terminal, when the power terminal is in the coverage range, θ a = 0.

[0135] Step 6: using the differential evolution algorithm of the hybrid simulated annealing algorithm to solve the unmanned aerial vehicle cooperative edge computing offloading model applied to the power terminal, to obtain the unmanned aerial vehicle cooperative edge computing offloading scheme applied to the power terminal.

[0136] The traditional differential evolution algorithm is easy to fall into local optimum, and when the optimization problem is relatively complex, the differential evolution algorithm also has the problems of easy premature, low global optimization efficiency and the like due to improper parameter setting. In order to solve these problems in the process of searching for the optimal solution, the application combines the simulated annealing algorithm. The differential evolution algorithm is divided into four steps of initialization, mutation, crossover and selection, and after the differential evolution algorithm performs the "selection" operation on an individual, the simulated annealing algorithm is executed on the individual again, the current population is disturbed, and the local optimum is avoided. Compared with the genetic algorithm (GA), the particle swarm algorithm and the like, the differential evolution algorithm (DE) has more advantages in convergence speed, robustness, precision and realizability.

[0137] The differential evolution algorithm combined with the simulated annealing algorithm is as follows:

[0138]

[0139]

[0140] The formula used by the differential evolution algorithm combined with the simulated annealing algorithm is as follows:

[0141] Formula (9): m i =x i +F·(x y -x j )

[0142] Formula (10):

[0143] Formula (11):

[0144] Formula (12):

[0145] Formula (13): Δ i =f(h i )-f(x i )

[0146] Formula (14):

[0147] The steps are as follows:

[0148] X 3N-dimensional vectors are randomly generated in the feasible region Z to form the first generation population;

[0149] Through mutation, the mutation individual m i of the i th individual x i in the population is calculated according to the scaling factor F;

[0150] Through crossover, the trial individual ni ;

[0151] By selection, determine its new offspring individual;

[0152] Form a new individual h by exchanging any two parts in the existing individual i ;

[0153] Calculate the fitness increment Delta i ;

[0154] Calculate the selection probability l i ;

[0155] Determine the value of individual x by the calculation probability of fitness increment i ;

[0156] Repeat the above steps until the maximum number of iterations is reached.

[0157] The application is directed to a wide range, sparse distribution of power equipment, communication difficulties and other complex environments, and proposes a UAV cooperative edge computing offloading scheme applied to power terminals. The scheme supports the use of power equipment through UAV cooperative edge computing offloading, and constructs a method that can meet the delay requirements of all power equipment while enabling high-priority tasks to be processed faster.

[0158] The application utilizes a fixed-wing UAV to carry an edge computing server as an edge node for dynamic deployment, which can cover complex environments and greatly improve the application range of the multi-access edge computing system. Through the priority-guided queue real-time scheduling algorithm, the sorting of power tasks in the waiting queue is flexibly adjusted, which can meet the delay requirements of all power equipment while enabling high-priority tasks to be processed faster. In view of the shortcoming of the differential evolution algorithm that it is easy to fall into local optimum, the algorithm is combined with the simulated annealing algorithm and priority adjustment, and finally better performance is obtained, achieving the purpose of minimizing the total delay and processing high-priority tasks first under the condition of meeting the delay requirements of all power equipment.

[0159] Embodiment two:

[0160] The embodiment of the application provides a UAV cooperative edge computing offloading system applied to power terminals, which comprises:

[0161] The acquisition module is used to acquire each power equipment parameter, UAV parameter and multi-access edge computing parameter.

[0162] Delay calculation module: for the parameters obtained, calculate the delay generated in each link between the multi-access edge computing system and the power equipment except the task queuing delay, including transmission delay d_sen, multi-access edge computing processing delay d_cal, power terminal processing delay d_equ and unmanned aerial vehicle relay delay d_re; the priority dominant queue real-time scheduling algorithm is used to adjust the sorting of the power task in the waiting queue, and the task queuing delay d_q is obtained;

[0163] First calculation module: for calculating the total time of processing data between the multi-access edge computing system and the power equipment based on the delay;

[0164] Model construction module: based on the shortest total time, construct the unmanned aerial vehicle cooperative edge computing offloading model applied to the power terminal;

[0165] Second calculation module: for solving the unmanned aerial vehicle cooperative edge computing offloading model applied to the power terminal by using the differential evolution algorithm of the hybrid simulated annealing algorithm, obtaining the unmanned aerial vehicle cooperative edge computing offloading scheme applied to the power terminal.

[0166] Embodiment three:

[0167] The embodiment of the application provides a computer device, comprising a processor and a storage medium;

[0168] The storage medium is used for storing instructions;

[0169] The processor is used for operating according to the instructions to perform the steps of the method of embodiment one.

[0170] Embodiment four:

[0171] The embodiment of the application further provides a computer readable storage medium, which stores a computer program, and the program is executed by a processor to realize the steps of the method of embodiment one.

[0172] Those skilled in the art should understand that the embodiments of the application can be provided as a method, a system or a computer program product. Therefore, the application can be in the form of a complete hardware embodiment, a complete software embodiment or an embodiment combining software and hardware aspects. Moreover, the application can be in the form of a computer program product implemented on one or more computer usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing computer usable program code.

[0173] The computer program instructions can also be loaded onto a computer or other programmable data processing apparatus to cause a series of operational steps to be performed on the computer or other programmable apparatus to produce a computer-implemented process such that the instructions which execute on the computer or other programmable apparatus provide steps for implementing the functions specified in the flowchart block or blocks. Figure 1 one or more flow or blocks. Figure 1 one or more flow or blocks.

[0174] These computer program instructions can also be stored in a computer readable memory that can direct a computer or other programmable data processing apparatus to function in a particular manner, such that the instructions stored in the computer readable memory produce an article of manufacture including instructions which implement the function specified in the flowchart block or blocks. Figure 1 one or more flow or blocks. Figure 1 one or more flow or blocks.

[0175] The computer program instructions can also be loaded onto a computer or other programmable data processing apparatus to cause a series of operational steps to be performed on the computer or other programmable apparatus to produce a computer-implemented process such that the instructions which execute on the computer or other programmable apparatus provide steps for implementing the functions specified in the flowchart block or blocks. Figure 1 one or more flow or blocks. Figure 1 one or more flow or blocks.

[0176] The above only is the preferred embodiment of the present application, it should be pointed out that, for those skilled in the technical field, without departing from the technical principles of the present application, can also make a number of improvements and variations, these improvements and variations should also be considered as the protection scope of the present application.

Claims

1. A method for UAV cooperative edge computing offloading applied to a power terminal, characterized in that, The application comprises the following steps: acquiring each power terminal parameter, unmanned aerial vehicle parameter and multi-access edge computing parameter; calculating the time delay of each link between the multi-access edge computing system and the power terminal except for the task queuing time delay based on the acquired parameters, including the transmission time delay d_sen, the multi-access edge computing processing time delay d_cal, the power terminal processing time delay d_equ and the unmanned aerial vehicle relay time delay d_re; wherein, the calculation of the time delay of each link between the multi-access edge computing system and the power terminal except for the task queuing time delay comprises: calculating the multi-access edge computing processing time delay d_cal by the following formula: In formula (1), g is the multi-access edge computing offloading ratio; D is the data collected by the power terminal; E_cal is the computing power of the unmanned aerial vehicle itself; calculating the power terminal processing time delay d_equ by the following formula: In formula (2), E_equ is the computing power of the power terminal itself; calculating the transmission time delay d_sen by the following formula: In formula (3), B is a bandwidth used by the power terminal; P is power; q is an antenna gain; w is a channel gain, w0 is a unit distance channel gain; l EU is a distance between the power terminal and the unmanned aerial vehicle which offloads data for the power terminal; σ 2 is a noise power; In the multi-access edge computing system composed of the unmanned aerial vehicle, the flight rule is a relative position fixed and equal distance, and the unmanned aerial vehicle flight trajectory is a clockwise circle, then the unmanned aerial vehicle relay time delay d_re is calculated by the following formula: In formula (4), R is the unmanned aerial vehicle flight radius; A is the number of unmanned aerial vehicles; a is the position sequence number of the unmanned aerial vehicle closest to the power terminal in the clockwise direction relative to the unmanned aerial vehicle transmitting the calculation result; v is the unmanned aerial vehicle flight speed; the maximum time delay d_wait that the power terminal needs to wait for the unmanned aerial vehicle to communicate with it is calculated by the following formula: In formula (5), r is the coverage radius of the unmanned aerial vehicle communication capability; adjusting the order of the power task in the waiting queue by using the priority dominant queue real-time scheduling algorithm, and obtaining the task queuing time delay d_q; calculating the total time for processing data between the multi-access edge computing system and the power terminal by the following formula: T_tal=max{d_equ,d_MEC} =max{d_equ,d_sen+d_q+d_cal+d_qr+d_re+d_wait+d_d} (6) In formula (6), T_tal is the total time for processing data between the multi-access edge computing system and the power terminal; d_equ is the power terminal processing time delay; d_sen is the transmission time delay; d_q is the task queuing time delay; d_cal is the multi-access edge computing processing time delay; d_qr is the queuing relay time delay; d_re is the unmanned aerial vehicle relay time delay; d_wait is the maximum time delay that the power terminal needs to wait for the unmanned aerial vehicle to communicate with it; d_d is the time delay of the data download; constructing an unmanned aerial vehicle cooperative edge computing offloading model applied to the power terminal based on the shortest total time, which is represented by the following formula: c.t.t n ≤E n 0≤b n ≤1 0≤β n ≤1 In formula (7), n is the power terminal serial number, N is the number of power terminals; t n is the total time for processing data between the multi-access edge computing system and the power terminal n, E n is the effective time for processing data between the multi-access edge computing system and the power terminal n; b n is the proportion of the channel occupied by the power terminal n; β n is the proportion of the computing resource used by the power terminal n; x a,n is the difference in the horizontal coordinate of the relative coordinates between the power terminal n and the unmanned aerial vehicle, y a,n is the difference in the vertical coordinate of the relative coordinates between the power terminal n and the unmanned aerial vehicle; g n is the offloading proportion of the power terminal n; D n is the data collected by the power terminal n; B0 is the downlink channel bandwidth; B1 is the uplink channel bandwidth; P n is the power of the power terminal n; q n is the antenna gain of the power terminal n; l n is the distance between the power terminal n and the unmanned aerial vehicle to which the data of the power terminal n is offloaded; θ a is the angle between the coverage range of the unmanned aerial vehicle closest to the power terminal and the power terminal, when the power terminal is in the coverage range, θ a = 0. solving the unmanned aerial vehicle cooperative edge computing offloading model applied to the power terminal by using the differential evolution algorithm of the hybrid simulated annealing algorithm, and obtaining the unmanned aerial vehicle cooperative edge computing offloading scheme applied to the power terminal. 2.The UAV cooperative edge computing offloading method applied to the power terminal of claim 1, wherein, The adjustment of the order of the power task in the waiting queue by using the priority dominant queue real-time scheduling algorithm comprises the following steps: If there is no task in the queue, the current task is directly added to the queue; if there is a task, the last task in the queue is sorted and replaced; If the current task is not in the same level or after the higher level task, but stops moving forward because it meets the time delay requirement, the replacement cycle ends; if it does not meet the time delay requirement, but the current task reaches the same level task and stops moving forward, the replacement cycle ends. The optimization step uses a differential evolution algorithm of a hybrid simulated annealing algorithm to solve the unmanned aerial vehicle cooperative edge computing offloading model applied to the power terminal, including: 3.The UAV cooperative edge computing offloading method applied to the power terminal of claim 1, wherein, Randomly generate X 3N-dimensional vectors in the feasible domain Z to form the first generation population; Determine the new offspring individuals by selection; By mutation, the mutation individual m of the i-th individual x in the population is calculated according to the scaling factor F i ; and i ; By crossover, using the ith individual and the mutated individual, according to the crossover probability cr, the experimental individual n is calculated i ; Repeat the above steps until the maximum number of iterations is reached. A new individual h is formed by exchanging any two parts of the existing individual i ; calculating the fitness increment Δ i ; Computing selection probability l i ; determining the probability of an individual x by the calculation of the fitness increment i the value of the number; It includes:

4. A UAV cooperative edge computing offloading system applied to a power terminal, characterized in that, An acquisition module is configured to acquire each power terminal parameter, unmanned aerial vehicle parameter, and multi-access edge computing parameter; A time delay calculation module is configured to calculate the time delay generated by each link between the multi-access edge computing system and the power terminal, including transmission time delay d_sen, multi-access edge computing processing time delay d_cal, power terminal processing time delay d_equ, and unmanned aerial vehicle relay time delay d_re, based on the acquired parameters, in addition to the task queuing time delay. The multi-access edge computing processing time delay d_cal is calculated by the following formula: In formula (1), g is the multi-access edge computing offloading ratio; D is the data collected by the power terminal; E_cal is the computing power of the unmanned aerial vehicle itself; The power terminal processing time delay d_equ is calculated by the following formula: In formula (2), E_equ is the computing power of the power terminal itself; The transmission time delay d_sen is calculated by the following formula: The unmanned aerial vehicle flies in a multi-access edge computing system, and the flight rules are relative position fixed and equal distance. The unmanned aerial vehicle flight trajectory is a clockwise circle. The unmanned aerial vehicle relay time delay d_re is calculated by the following formula: In Equation (3), B is a bandwidth used by the power terminal; P is power; q is an antenna gain; w is a channel gain, w0is a unit distance channel gain; l EU is a distance between the power terminal and the drone which offloads data for the power terminal; σ 2 is a noise power; In formula (4), R is the flight radius of the unmanned aerial vehicle; A is the number of unmanned aerial vehicles; a is the position number of the unmanned aerial vehicle closest to the power terminal in the clockwise direction relative to the unmanned aerial vehicle transmitting the calculation result; v is the flight speed of the unmanned aerial vehicle; The maximum time delay d_wait that the power terminal needs to wait for the unmanned aerial vehicle to communicate with it is calculated by the following formula: In formula (5), r is the coverage radius of the communication capability of the unmanned aerial vehicle; A priority-based queue real-time scheduling algorithm is used to adjust the order of the power task in the waiting queue, and the task queuing time delay d_q is obtained; A first calculation module is configured to calculate the total time for the multi-access edge computing system and the power terminal to process data by the following formula: T_tal=max{d_equ,d_MEC} ​ = max{d_equ, d_sen + d_q + d_cal + d_qr + d_re + d_wait + d_d} (6) In formula (6), T_tal is total time for processing data between the multi-access edge computing system and the power terminal; d_equ is power terminal processing delay; d_sen is transmission delay; d_q is task queuing delay; d_cal is multi-access edge computing processing delay; d_qr is queuing relay delay; d_re is unmanned aerial vehicle relay delay; d_wait is maximum delay that the power terminal needs to wait until the unmanned aerial vehicle can communicate with it; and d_d is delay of downlink data. A model construction module: based on the shortest total time, an unmanned aerial vehicle cooperative edge computing offloading model applied to the power terminal is constructed, and the model is represented by the following formula: c.t.t n ≤E n 0≤b n ≤1 0≤β n ≤1 In formula (7), n is the power terminal serial number, N is the number of power terminals; t n is the total time for processing data between the multi-access edge computing system and the power terminal n, E n is the effective time for processing data between the multi-access edge computing system and the power terminal n; b n is the proportion of the channel occupied by the power terminal n; β n is the proportion of the computing resource used by the power terminal n; x a,n is the difference between the horizontal coordinate of the relative coordinate of the power terminal n and the unmanned aerial vehicle, y a,n is the difference between the vertical coordinate of the relative coordinate of the power terminal n and the unmanned aerial vehicle; g n is the offloading proportion of the power terminal n; D n is the data collected by the power terminal n; B0 is the downlink channel bandwidth; B1 is the uplink channel bandwidth; P n is the power of the power terminal n; q n is the antenna gain of the power terminal n; l n is the distance between the power terminal n and the unmanned aerial vehicle to which the data of the power terminal n is offloaded; θ a is the angle between the coverage range of the unmanned aerial vehicle closest to the power terminal and the power terminal, when the power terminal is in the coverage range, θ a = 0. A second calculation module: used for solving the unmanned aerial vehicle cooperative edge computing offloading model applied to the power terminal by using a differential evolution algorithm of a hybrid simulated annealing algorithm, to obtain an unmanned aerial vehicle cooperative edge computing offloading scheme applied to the power terminal.

5. A computer device, comprising: comprising a processor and a storage medium; the storage medium is used to store instructions; the processor is used to operate according to the instructions to perform the steps of the method according to any one of claims 1-3.

6. A computer readable storage medium having stored thereon a computer program, characterized in that The program is executed by the processor to implement the steps of the method according to any one of claims 1-3.

Images