Research And Implementation Of Paticle Swarm Optimization Based On Membrane Computing In Cloud Resource Scheduling

Cloud computing is the development of grid computing.For the cloud system,how to scientifically and reasonably perform cloud computing resource scheduling is more than important.This Thesis proposes an improved particle swarm strategy based on membrane computing that is applied to resource scheduling in cloud computing.The main contributions are as follows:1.An improved particle swarm optimization algorithm based on membrane computing is proposed.The particle swarm intelligence algorithm is combined with membrane computing,and divided the membrane system into the main membrane and the auxiliary membrane.It is improved according to certain rules in the main membrane and the auxiliary membrane respectively.Different improved particle swarm algorithms were used according to different duties between the main film and the auxiliary film.Experimental results show that this improvement effectively improves the robustness of the algorithm.2.A CLPSO algorithm is proposed.Based on the characteristics of membrane computing,the intelligent algorithm in the auxiliary membrane requires a relatively strong global search capability and racial diversity.Therefore,In the thesis,the idea of neighborhood is introduced on the basis of chaotic thoughts.That is,not only the change of chaotic sequence but also the state of adjacent particles should be considered in the iterative update process of particles in order to achieve a better global search effect.3.An improved particle swarm optimization algorithm that combines fast convergence and multi-scale variation escape is proposed.Based on the characteristics of membrane computing,the main membrane requires a relatively strong local search capability and accelerates the convergence rate.Therefore,the idea of the improved algorithm proposed in the thesis is to perform multiple local search on high-quality particles so that they can better disseminate high-quality information.On this basis,the idea of multi-scale adaptive mutation escape was proposed.According to the calculation of the variance of different scales,whether the premature escaping was selected or not.Experimental results show that this improvement improves the accuracy of local search.4.According to the characteristics of cloud resource scheduling and physical model,a mathematical model of cloud resource scheduling is abstracted.The improvedalgorithm is fully integrated with the cloud resource scheduling model to propose a feasible strategy for cloud resource scheduling.CloudSim was introduced.The platform was recompiled and the simulation experiments of the proposed algorithm were performed.The simulation results were compared with other commonly used algorithms.Theoretical analysis and experiments illustrate that the M-PSO algorithm has certain advantages in terms of energy consumption,SLA violation rate and performance,and can achieve effective results.