Research On Task Scheduling Strategy Based On Improved Particle Swarm Algorithm In The Cloud Environment

With generation of the large amounts of data,human beings gradually entered the era of big data that then produces a series of new things.Cloud computing is one of them.The emergence and development of cloud computing updates new approach of a lot of data,which makes data processing more efficient and convenient.Cloud computing makes the computer hardware and software as the resources that provide basic services to human beings,which are just like the social public infrastructure.However,reasonable scheduling of these resources has become a problem of cloud computing data center.Task scheduling of cloud data processing center is the key to the whole calculation process.Based on cloud computing data center,this thesis,treating particle swarm optimization algorithm as the research object,aims to conduct an in-depth study of particle swarm optimization algorithm,and to improve the particle swarm optimization algorithm.The improved algorithm is applied to the task scheduling in the cloud environment.Because standard particle swarm optimization is randomly generated when initializing particles that can't ensure that particles uniformly distribute in the search space,the thesis implements the chaotic sequence set for initial position and velocity of particle individuals at the initialization period,which can make the initial particles evenly distribute in the solution space.Considering that particle will appear and its particle diversity will decrease in the late iterations etc,the author adds the concept of particle gravity to updating formula of particle velocity in order to enhance the global optimization ability of particles.Finally,all of these will be tested by the CloudSim simulator.The experimental results show that the algorithm can get better scheduling results and faster convergence speed.In the process of solving practical problems,most problems need to be taken a number of objectives into consideration,optimized and weighed its multiple targets.This thesis conducts an in-depth study of multi-objective particle swarm optimization algorithm and applied it to multi-objective task scheduling in cloud environment.In the actual application of cloud environment,in addition to the total completion time of the task,the total cost of the task is also one of the key factors needed to be considered.Two objectives,the total completion time and the total completion cost are studied.The experimental results show that a set of optimal solution sets can be obtained for scheduling selection through the multi-objective particle swarm optimization algorithm.