Research On The Application Of Cloud Resource Scheduling By Dynamic Evolutionary Algorithm

Dynamic multi-objective optimization evolutionary algorithm is widely used in engineering field to deal with a class of time-related dynamic multi-objective optimization problems.The efficiency of cloud resource scheduling is the key factor that affects the efficiency of cloud computing,and how to improve the efficiency of cloud resource scheduling is a problem that is concerned in cloud computing field.Many dynamic multi-objective optimization problems have been found in cloud resource scheduling.There are many advantages to use multi-objective optimization evolutionary algorithm in dealing with multi-objective optimization problems.It has become a hot research direction in the field of cloud computing research.This paper focuses on the application of dynamic multi-objective optimization evolutionary algorithm in cloud resource scheduling.It firstly introduces the characteristics and research status of cloud resource scheduling.Then,it analyzes the advantages and disadvantages of the multi-objective optimization evolutionary algorithms in the treatment of the cloud resource scheduling problem.Based on historical information of resource schedule,a novel dynamic multi-objective optimization evolutionary algorithm is proposed to resolve the cloud resource scheduling problem.The main contributions of this paper include the following three aspects:Firstly,this paper proposes a multi-objective optimization cloud resource scheduling model.This model is implemented by means of virtual machine migration which involves energy consumption of cloud,cloud service ability after migration and migration cost by virtual machine migration.It is a balance resource scheduling model.Secondly,a new coding strategy is designed to represent the cloud environment.So it takes into account by several principal principles of coding strategy: non-redundancy,legality and completeness.This coding strategy is very simple and one-toone,with no repetitive coding problems.Then,we use this coding strategy to improve the multi-objective evolutionary algorithm and apply it to cloud resource scheduling.Experiments show that the proposed multi-objective optimization evolutionary algorithm is high efficiency in cloud resource scheduling.Finally,this paper proposes a dynamic multi-objective optimization algorithm based on historical information and historical information prediction,and applies it to dynamic cloud resource scheduling.Combining with the characteristics of the dynamic change of cloud resource scheduling,the multi-objective evolutionary algorithm is improved.By using historical information in evolutionary process to predict the trend of population change direction,the diversity rate of the algorithm is accelerated.At the same time,it makes full use of the historical information in the process of evolution to ensure the convergence of the population.Through a large number of experiments,its performance in dynamic cloud resource scheduling is verified.