基于PSO-LSTM模型的手足口病发病趋势预测研究

打开文本图片集

中图分类号：TP183；TP39 文献标识码：A 文章编号：2096-4706（2025）19-0168-05

Abstract： This paperanalyzes the incidence trendof Hand，Foot，and MouthDisease（HFMD）inEnshi Prefecture，and discusses the application effectof thecombined model of Particle Swarm Optimization （PSO）andLong Short-Term Memory （LSTM） in predicting the incidence rateof HFMD in EnshiPrefecture.ThePSO-LSTMmodel isconstructed andfitted using weekly HFMDincidencedata fromJanuary2010 to June 2022 inEnshiPrefecture.The model isusedtopredictthe incidence trendfortheext16months，anditsperformanceiscomparedwithtraditionalLST，SupportVectorRegresson（SV）nd Random Forest （RF）models.The PSO-LSTM model achieves superior performance with Mean Squared Error （MSE），Mean Absolute Error （MAE），and coefficient of determination （R2） values of1203.23，16.36，and 0.80，respectively，outperforming the traditionalLSTM，SVR，andRFmodels.The PSO-LSTMmodel demonstrates higher predictiveaccuracyin predicting HFMD incidence trends，significantlyenhancingepidemic situation warningand preventionandcontrol effciency，and providing more precise decision support for disease prevention and control.

Keywords：Hand，FootandMouthDisease;ParticleSwarmOptimization;LongShort-TermMemory;SupportVector Regression; Random Forest; infectious disease prediction

0 引言

手足口病（HandFootMouthDisease，HFMD）是一种主要由肠道病毒71型（EV71）和柯萨奇A组16型（CoxA16）引起的高发于5岁以下儿童的传染性疾病，其典型症状包括发热和手、足、口腔等部位出现散发性皮疹和疱疹。（剩余5635字）