摘要
针对物联网网络节点流量预测的挑战,提出了基于改进蜂群算法的快速预测方法。该方法通过计算初始流量时序长度、构建流量时序权重矩阵,并引入全局最优解和自适应惯性参数的改进蜂群算法。实验结果显示,该方法在MAE、RMSE上表现最优(分别为65.2490和73.6583),且R2值最接近1(0.9969),验证了其预测的高准确性。该方法有效应对了网络流量动态性和复杂性,提升了预测精度。
A fast prediction method based on an improved bee colony algorithm is proposed to address the challenge of predicting node traffic in the Internet of Things network.This method calculates the initial flow time series length,constructs a flow time series weight matrix,and introduces an improved bee colony algorithm with global optimal solution and adaptive inertia parameters.The experimental results showed that the method performed the best on MAE and RMSE(65.2490 and 73.6583,respectively),and the R2 value was closest to 1(0.9969),verifying its high accuracy in prediction.This method effectively addresses the dynamic and complex nature of network traffic,improving prediction accuracy.
作者
王星星
魏晓梦
WANG Xingxing;WEI Xiaomeng(Henan Mechanical&Electrical Vocational College,Henan Zhengzhou 451100,China)
出处
《长江信息通信》
2025年第2期46-48,共3页
Changjiang Information & Communications
关键词
改进蜂群算法
物联网
网络节点
预测方法
权重矩阵
improving swarm algorithm
Internet of things
network node
prediction method
weight matrix
