CN-115437771-B - Online scheduling method for power tasks of power distribution internet of things equipment

CN115437771BCN 115437771 BCN115437771 BCN 115437771BCN-115437771-B

Abstract

An online dispatching method for power tasks of power distribution internet of things equipment belongs to the field of edge computing of the power distribution internet of things. The method is characterized by comprising the following steps of 1, determining the number of edge devices and cloud ends, creating containers and devices, performing communication setting of the edge devices and the cloud ends, 2, creating tasks and scheduling lists, 3, building a container-micro service mapping relation, 4, determining earliest micro service execution time, 5, distributing tasks and micro services thereof when edge computing system resources are not limited and edge computing system resources are limited, and 6, deleting micro service information from the scheduling list after the micro services are distributed and executed. In the power distribution internet of things equipment power task online scheduling method, different tasks are scheduled according to different resource conditions of a system, more tasks can be processed in a shorter time, and the completion rate of important tasks when the system resources are limited is ensured.

Inventors

CHEN YU
CHENG QIAN
DING RUI
WANG WEI
ZHANG XINHUI
PENG KE

Assignees

山东理工大学

Dates

Publication Date: 20260512
Application Date: 20220916

Claims (5)

1. The power task online dispatching method for the power distribution internet of things equipment is characterized by comprising the following steps of: step 1, determining the quantity of edge devices and cloud ends, creating containers and devices, and performing communication setting of the edge devices and the cloud ends; Step 2, creating a task and a scheduling list; Step 3, establishing a mapping relation between the container and the micro service; Step 4, determining the earliest execution time of the micro service according to the equipment capability constraint and the micro service priority constraint; Step 5, according to the earliest execution time of the micro-service determined in the step 4, respectively distributing the task and the micro-service when the edge computing system resources are not limited and the edge computing system resources are limited; step 6, deleting the information of the micro service from the dispatch list after the micro service is distributed and executed; the step 4 comprises the following steps: step 4-1, latency under device capability constraints: Wherein: representing micro services At the earliest end time of device e n , micro-services Is the wait queue for container c i in device e n Tail end microservice of (a), and WQ n represents the set of tail micro-services for which all containers of device e n wait for a queue; step 4-2, earliest execution time under priority constraint; the micro-service needs to be executed after the data of its precursor micro-service is transferred to, and therefore: Wherein, the Is that At the earliest execution time of device e n , traversing the edge devices and cloud may get the minimum of this time, Is a micro-service The device to be allocated is a device to be allocated, Step 4-3, determining micro-services The earliest execution time in device e n is: The end time is: ; in step 5, when the edge computing system resource is not limited to allocate the task and the micro-service thereof, the method comprises the following steps: Step 5-1, calculating earliest execution time of the latest micro-service of each scheduling list according to the earliest execution time of the micro-service determined in the step 4-3, selecting the micro-service with the earliest execution time, if a new task arrives in the waiting time, discarding the scheduling, and only updating the current time, the remaining time of each container and the container state; Step 5-2, if no new task arrives within the waiting time, distributing the micro services by utilizing a micro service distribution formula; and 5-3, distributing the micro services to the rest micro services in sequence according to a micro service distribution formula.
2. The method for online dispatching of power tasks of the power distribution internet of things equipment according to claim 1, wherein the step 2 comprises the following steps: step 2-1, creating a task model; modeling a plurality of dependent microservices using a directed acyclic graph, denoted G m ={V m ,E m ,W m , Wherein: Is a node of the graph, represents k micro-services of task call, E m represents a set of directed edges, represents dependency relationships between micro-services, and W m represents a set of weights of each directed edge; Step 2-2, modifying a task topology structure; Step 2-3, determining task information; the set of power tasks is denoted s= { S 1 ,s 2 ,s 3 ,…,s m ,..and task S m is denoted S m ={a tm ,d tm ,l m ,G m , Wherein a tm represents the arrival time of the task s m , d tm represents the deadline of the task s m , and l m represents the value of the task s m type for distinguishing the conventional task, the alarm task and the fault processing task; step 2-4, estimating the execution time of each micro service in each device according to the history information; step 2-5, calculating task priority through fuzzy logic; Step 2-6, traversing each node from the virtual export micro-service forward, and obtaining the micro-service scheduling priority in a recursion mode; And 2-7, creating a scheduling list.
3. The online dispatching method for the electric power tasks of the power distribution internet of things equipment according to claim 2 is characterized by comprising the following steps of: Step 2-5-1, adopting a triangle membership function to carry out fuzzification; step 2-5-2, reasoning by using fuzzy rules; Step 2-5-3, using centroid method to perform the defuzzification calculation.
4. The method for online scheduling of power tasks of power distribution thing-connected equipment according to claim 1, wherein in the step 5, when the tasks and the micro-services thereof are allocated by limiting the resources of the edge computing system, it is assumed that a scheduling sequence with the highest priority in the scheduling list D is T m , when the priority of T m is greater than 0.8, the micro-service with the forefront in the scheduling sequence T m is allocated to the target equipment according to a micro-service allocation formula, and if a plurality of tasks with the same priority are allocated, one scheduling is randomly selected.
5. The online dispatching method for the electric power tasks of the power distribution internet of things equipment according to claim 1 or 4, wherein the micro-service distribution formula is as follows: Wherein, the Is a micro-service The device to be allocated is a device to be allocated, Is that At the earliest execution time of device e n .

Description

Online scheduling method for power tasks of power distribution internet of things equipment Technical Field An online dispatching method for power tasks of power distribution internet of things equipment belongs to the field of edge computing of the power distribution internet of things. Background Aiming at the new requirements of the diversification, the dynamics and the differentiation of the tasks of the distribution area and the service range to the deep user side, the side equipment forms a new architecture taking micro-services as an execution carrier. Due to the advantages of good isolation performance and light weight of the container, the application programs of the micro service type are deployed through the container, the processing of the edge equipment on multiple tasks is realized, and task isolation and data security are ensured. The trend in the development of edge devices is to generalize, i.e. to run different types of applications on the same hardware system. However, the edge device is a resource-limited device, only a limited number of service containers can be supported, and the application is increased along with the expansion of new services, so that the contradiction between the diversified demands of the services and the limited device resources is increasingly prominent. Scheduling tasks is required to solve the above problems. The task scheduling process refers to distributing task requests from the terminal to the side equipment or the cloud end, providing corresponding resources and executing tasks in a cooperative mode. The core is to make reasonable scheduling decisions according to the system conditions and task characteristics. The micro-service architecture and the container technology enable the application architecture and the resource deployment mode of the cloud edge of the power distribution Internet of things to be changed, and further the scheduling processing of tasks is changed. Specifically, first, the task is composed of a plurality of micro services with dependency relationship, and granularity is finer, so that scheduling and execution sequence of each micro service are constrained. Second, each micro-service may correspond to a container to process it, with the running state of the container being dynamically changed. Finally, the side equipment and the cloud bear a plurality of kinds of service containers at the same time, the information, the functions and the requirements of various requests are greatly different, the single task scheduling sequence and the single task scheduling mode cannot meet the diversified requirements of the tasks, and the tasks are required to be distributed comprehensively according to the task characteristics. In the prior art, the influence of the power task characteristics and the container state on the scheduling decision under the micro-service architecture is not considered, the task is scheduled in an offline mode, and the research on the online scheduling of multiple tasks is lacking. Therefore, the technical scheme of online scheduling research for reducing the response time delay of tasks and improving the task completion rate is designed and becomes a problem to be solved in the field. Disclosure of Invention The invention aims to solve the technical problems of overcoming the defects of the prior art, and providing the power task online scheduling method for the power distribution internet of things equipment, which is used for scheduling different tasks according to different resource conditions of a system, can process more tasks in a shorter time and ensures the completion rate of important tasks when the system resources are limited. The technical scheme adopted by the invention for solving the technical problems is characterized by comprising the following steps: step 1, determining the quantity of edge devices and cloud ends, creating containers and devices, and performing communication setting of the edge devices and the cloud ends; Step 2, creating a task and a scheduling list; Step 3, establishing a mapping relation between the container and the micro service; Step 4, determining the earliest execution time of the micro service according to the equipment capability constraint and the micro service priority constraint; Step 5, according to the earliest execution time of the micro-service determined in the step 4, respectively distributing the task and the micro-service when the edge computing system resources are not limited and the edge computing system resources are limited; And 6, deleting the information of the micro service from the scheduling list after the micro service is allocated and executed. Preferably, the step 2 includes the following steps: step 2-1, creating a task model; Modeling a plurality of dependent microservices using a directed acyclic graph, denoted G m＝{Vm,Em,Wm, Wherein: Is a node of the graph, represents k micro-services of task call, E m represents a set of directed edges, represen