基于改进极限学习机的应急通信网络多重拥堵风险预警方法  

Multiple Congestion Risk Warning Method for Emergency Communication Networks Based on Improved Extreme Learning Machine

作　　者：钱立遥 QIAN Liyao(Zhejiang Provincial Emergency Management Digital and Technology Center,Hangzhou,Zhejiang 310000,China)

机构地区：[1]浙江省应急管理数字与技术中心,浙江杭州310000

出　　处：《计算技术与自动化》2024年第4期179-184,共6页Computing Technology and Automation

摘　　要：对风险数据特征提取不准确,导致预警精度低,因此,提出了基于改进极限学习机的应急通信网络多重拥堵风险预警方法。采集与标准化处理数据,实现对风险数据的预处理,在改进极限学习机的支持下,进行特征差和特征向量权重的计算,完成对风险数据的特征提取,通过对不同数据进行加权处理,构建风险预警模型,通过计算通信网络的风险值划分风险预警等级,实现通信网络风险预警。通过上述设计,完成对应急通信网络多重拥堵风险预警方法的设计。在仿真实验中,与以往的应急通信网络多重拥堵风险预警方法相比,设计的基于改进极限学习机的应急通信网络多重拥堵风险预警方法预警精度为100%,对风险的预警更准确。Inaccurate feature extraction of risk data leads to low warning accuracy.Therefore,a multiple congestion risk warning method for emergency communication networks based on improved extreme learning machine is proposed.Collect and standardize data to preprocess risk data.With the support of improved extreme learning machine,calculate feature differences and feature vector weights to complete feature extraction of risk data.By weighting different data,construct a risk warning model,calculate the risk value of the communication network,classify the risk warning level,and achieve communication network risk warning.Through the above processes,the design of multiple congestion risk warning methods for emergency communication networks is completed.In the simulation experiment,compared with previous emergency communication network multiple congestion risk warning methods,the designed multiple congestion risk warning method for emergency communication network based on improved extreme learning machine has a warning accuracy of 100%,and is more accurate in risks warning.

关 键 词：改进极限学习机 应急通信网络 多重拥堵 风险预警 

分 类 号：G642[文化科学—高等教育学]