基于多策略粒子群优化RBF的云资源预测模型  

Cloud resource prediction model based on multi-strategy particle swarm optimization RBF

作　　者：杨迪[1] 刘思源 王鹏[1] 杨华民[1] RBF YANG Di;LIU Si-yuan;WANG Peng;YANG Hua-min(School of Computer Science and Technology,Changchun University of Science and Technology,Changchun 130000,China)

机构地区：[1]长春理工大学计算机科学技术学院,吉林长春130000

出　　处：《计算机工程与设计》2023年第4期1073-1080,共8页Computer Engineering and Design

基　　金：中央引导地方科技发展基金项目(202002038JC)。

摘　　要：针对云计算资源利用率低等问题,构建基于多策略粒子群优化RBF神经网络的云资源预测模型(MPSO-RBF)。采用改进的粒子群算法对RBF神经网络参数进行优化,避免随机初始化参数引起的预测精度低等问题;对于粒子群容易陷入局部最优解等问题,采用动态惯性权重、自适应学习因子和变异粒子位置3种策略对粒子群进行改进,提高算法的寻优能力。基于云计算资源负载数据,将该模型与BP、RBF和PSO-RBF模型进行对比实验,验证了该模型具有良好的性能。Aiming at the problem of low utilization of cloud computing resources,a cloud resource prediction model based on multi-strategy particle swarm optimization RBF neural network(MPSO-RBF)was constructed.The improved particle swarm optimization algorithm was used to optimize the parameters of RBF neural network,which effectively avoided the problem of low prediction accuracy caused by random initialization parameters.For the problem that particle swarm optimization is easy to fall into local optimal solution,three strategies of dynamic inertia weight,adaptive learning factor and mutation particle position were used to improve the particle swarm optimization,which improved the optimization ability of the algorithm.Based on cloud computing resource load data,the model was compared with BP,RBF and PSO-RBF models to verify the better performance of the proposed model.

关 键 词：云计算 负载资源预测 粒子群算法 径向基神经网络 柯西分布 学习因子 惯性权重 

分 类 号：TP391.3[自动化与计算机技术—计算机应用技术]