Research On Task Scheduling Based On Improved Particle Swarm Optimization Algorithm In Cloud Computing Environment

Cloud computing,as a new business service model,has been widely used in different industries.For example,in the financial sector,cloud computing can help banks and securities companies better manage large amounts of data;in the healthcare sector,cloud computing can support medical institutions for remote consultations,research and treatment.However,with the rapid development of cloud computing,there is a growing concern about energy consumption,which limits the overall performance of cloud data centers in cloud computing.With energy saving,it is particularly important to efficiently allocate resource nodes for users’ task requirements in cloud computing.Existing task scheduling algorithms can hardly meet the demand,and there is an urgent need for better task scheduling algorithms to improve cluster efficiency and reduce energy consumption.Traditional Particle Swarm Optimization(PSO)algorithm is widely used to solve the task scheduling problem in cloud computing,but its convergence speed is fast and accuracy is low,and the cluster energy consumption problem is easily ignored.In order to solve this problem,this thesis focuses on cloud computing resource consumption and task scheduling strategy.To address the problem that the energy efficiency ratio of cloud computing task scheduling algorithms is not high,opposition-based learning and chaotic mapping mechanisms are introduced into the standard particle swarm algorithm to make up for the shortcomings of existing particle swarm task scheduling algorithms.The main research work of this thesis includes:(1)A chaotic mapping Adaptive Particle Swarm Optimization algorithm based on Oppositionbased learning(OAPSO)is proposed.The algorithm can make the initial population of particles more uniformly distributed in the initial solution space,change the optimal-seeking ability of particles,and improve the ability of the algorithm to jump out from the local optimum.The results of multiple sets of simulation experiments on the proposed algorithm in Cloudsim,a cloud computing simulation platform,show that the OAPSO algorithm has more efficient energy saving efficiency,which is 29.7%,28.4%,and 16.1% higher compared to several other algorithms.(2)The OAPSO algorithm can improve the resource utilization of the system,but the adopted task scheduling energy consumption model considers the completion time and energy consumption of virtual machines to be equivalent and is not suitable for solving largescale task scheduling problems.Based on the OAPSO algorithm,an Improved chaotic mapping Adaptive Particle Swarm Optimization algorithm based on Opposition-based learning(IOAPSO)is proposed,and the evaluation metrics of the algorithm are also enriched.The proposed algorithm is verified on Cloudsim,and the experimental results show that the overall performance of the algorithm outperforms other control algorithms and has more significant advantages in the scheduling problem of large-scale tasks.In addition,the algorithm has higher resource utilization and better energy efficiency compared with other control algorithms.By deploying the above algorithm,it effectively shortens the task completion time when large-scale tasks are invoked in the cloud computing platform and saves the purchase cost of cloud computing resources.