摘要
准确预测云资源短期负载对提高云平台资源管理效率、保障云服务质量至关重要。针对传统模型在面对小样本、非线性云资源负载数据时预测精度不高,提出一种基于变分模态分解(VMD)和改进麻雀搜索算法(ISSA)优化最小二乘支持向量机(LSSVM)的云资源短期负载预测模型。将原始负载数据通过VMD分解成多个相对平稳的模态分量;对麻雀搜索算法进行优化,增强种群多样性,提高寻优性能和收敛速度。利用改进麻雀搜索算法优化LSSVM的关键参数,建立VMD-ISSA-LSSVM预测模型。利用Wikipedia网站的云资源负载数据进行仿真,结果表明,所提模型在预测精度上优于参照模型。
Accurate prediction of short-term cloud resource load is crucial to improve the efficiency of cloud platform resource management and guarantee the quality of cloud services.Traditional prediction models have low prediction accuracy in facing small samples and nonlinear cloud resource load data,hence,a cloud resource short-term load prediction model based on variational modal decomposition(VMD)and improved sparrow search algorithm(ISSA)optimized least squares support vector machine(LSSVM)is proposed.Firstly,the original data are decomposed into multiple smooth modal components by VMD.Secondly,the sparrow search algorithm is optimized to enhance its population diversity and improve its optimal search performance and convergence speed.Finally,the key parameters of LSSVM are optimized using the improved sparrow search algorithm to establish the VMD-ISSA-LSSVM prediction model.Simulations are performed using cloud resource load data from Wikipedia Website,and the experimental results show that the model proposed in this paper outperforms the reference model in terms of prediction accuracy.
作者
杨哲兴
谢晓兰
李水旺
YANG Zhexing;XIE Xiaolan;LI Shuiwang(College of Information Science and Engineering,Guilin University of Technology,Guilin 541004,Guangxi,China;Guangxi Key Laboratory of Embedded Technology and Intelligent Systems,Guilin University of Technology,Guilin 541004,Guangxi,China)
出处
《实验室研究与探索》
CAS
北大核心
2023年第6期117-124,共8页
Research and Exploration In Laboratory
基金
国家自然科学基金项目(62262011)
广西自然科学基金项目(2021JJA170130)。
关键词
云计算
负载预测
麻雀搜索算法
变分模态算法
最小二乘支持向量机
cloud computing
load prediction
sparrow search algorithm
variational modal algorithm
least squares support vector machine
