Particle swarm optimization workflow scheduling method based on directional search
A technology of particle swarm optimization and directional search, applied in the directions of program startup/switching, resource allocation, program control design, etc., can solve problems such as low correctness, local optimum, complex description, etc., and reduce scheduling costs and possibilities , good search speed effect
Pending Publication Date: 2021-12-17
HANGZHOU DIANZI UNIV
View PDF0 Cites 0 Cited by
- Summary
- Abstract
- Description
- Claims
- Application Information
AI Technical Summary
Problems solved by technology
Classical algorithms such as first-come-first-serve and HETF algorithms are easy to implement and work efficiently, but low correctness, complex descriptions, and simple scheduling strategies all lead to poor performance
Therefore, the classical algorithm does not achieve good results in workflow scheduling in a distributed environment.
As an intelligent algorithm, the particle swarm optimization algorithm has the advantages of few parameters, simple algorithm, and fast convergence speed, and is widely used in workflow scheduling. and other shortcomings
Method used
the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
View moreImage
Smart Image Click on the blue labels to locate them in the text.
Smart ImageViewing Examples
Examples
Experimental program
Comparison scheme
Effect test
Embodiment Construction
[0090] A particle swarm optimization workflow scheduling method based on directional search in the present invention will be further described below in conjunction with the accompanying drawings.
[0091] see figure 1 , shown as a flow chart of the method, which specifically includes the following steps:
[0092] Step S1: Define the workflow task constraint relationship in the edge environment
[0093] The task structure of the workflow is represented by a directed acyclic graph DAG; DAG is expressed as G=(T, E), where T={t 1 ,t 2 ,...,t n} represents n tasks that make up the workflow, task t i The computational workload is w i ;E={e ij | e ij =i ,t j >∧(t i ,t j )∈T×T,i≠j} represents the set of time or communication constraints between tasks; where e ij Indicates the task t i for task t j The predecessor task, task t j for task t i the successor task of ; task t i The set of predecessor tasks and the set of successor tasks are denoted as pre(t i ) and suc(t ...
the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
Login to View More PUM
Login to View More Abstract
The invention discloses a particle swarm optimization workflow scheduling method based on directional search. Non-linear inertia weight is adopted in the particle swarm algorithm, selection and mutation operation are carried out through a direction search process, and global optimization and local search capability of the algorithm is enhanced. Meanwhile, the scheduling problem of the workflow application and the objective function based on the two optimization factors are definitely formalized, and the optimal scheme of workflow scheduling can be found under the condition that task processing delay and cost are balanced. On the basis of a nonlinear inertia weight method, the global and local search capabilities of particles can be balanced and adjusted in different stages of iteration of the particle swarm optimization; and a directional search process is added when the particles are updated, so that the particles can escape from the interference of a local extreme value, the diversity of a particle swarm is kept, the possibility that the particles fall into local optimum is reduced, and a workflow scheduling scheme conforming to expectation can be generated.
Description
technical field [0001] The invention belongs to the field of workflow scheduling under a mobile edge computing environment, and in particular relates to a particle swarm optimization workflow scheduling method based on directional search. Background technique [0002] With the increasing popularity of the Internet of Things, edge computing has become a key driver to provide computing resources, storage and network services closer to the edge on the basis of cloud computing. Workflow scheduling in this distributed environment is considered to be an NP-hard problem, and existing methods may not be suitable for task scheduling with multiple optimization objectives in complex applications. [0003] The current workflow scheduling algorithms are mainly divided into two categories: classic algorithms and intelligent algorithms. Classical algorithms such as first-come-first-served and HETF algorithms are easy to implement and have high work efficiency, but low correctness, complex...
Claims
the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
Login to View More Application Information
Patent Timeline
Login to View More Patent Type & Authority Applications(China)
IPC IPC(8): G06F9/48G06F9/50G06N3/00
CPCG06F9/4843G06F9/5072G06N3/006
Inventor 袁友伟吴浩天钱逯彭瀚陆炎迪门中秀鄢腊梅
Owner HANGZHOU DIANZI UNIV