Chinese Journal of Vector Biology and Control ›› 2021, Vol. 32 ›› Issue (6): 744-748.DOI: 10.11853/j.issn.1003.8280.2021.06.018

• Technology and Method • Previous Articles     Next Articles

Application of autoregressive integrated moving average model in predicting the trend of rodent density in Shandong province, China

SUN Qin-tong, HAN Ying-nan, LIU Yan, LAI Shi-hong, WANG Xue-jun, KANG Dian-min   

  1. Institute for Disinfection & Vector Control, Shandong Center for Disease Control and Prevention, Ji'nan, Shandong 250014, China
  • Received:2021-05-12 Online:2021-12-20 Published:2021-12-15
  • Supported by:
    Supported by the National Major Science and Technology Project of China (No. 2017ZX100303404) and the Medical Health Science and Technology Program of Shandong Province (No. 202012051157)

应用自回归移动平均(ARIMA)模型预测山东省鼠密度趋势

孙钦同, 韩英男, 刘言, 赖世宏, 王学军, 康殿民   

  1. 山东省疾病预防控制中心消毒与病媒生物防制所, 山东 济南 250014
  • 通讯作者: 王学军,E-mail:bmfzs@126.com
  • 作者简介:孙钦同,男,硕士,主管医师,主要从事媒介生物监测及控制工作,E-mail:sunqtsdcdc@163.com
  • 基金资助:
    国家科技重大专项(2017ZX100303404);山东省医药卫生科技发展计划(202012051157)

Abstract: Objective To investigate the density of rodents at the national surveillance sites (Ji'nan, Qingdao, and Liaocheng) in Shandong province of China and to discuss the application of autoregressive integrated moving average (ARIMA) model in rodent density prediction and warning. Methods The night trap/cage method was used to monitor the density of rodents, and R 3.6.2 software was used to establish the ARIMA model for rodent density data in each monitoring month from 2010 to 2018. The predicted rodent density was compared with the actual rodent density in each monitoring month in 2019 to evaluate the prediction effect of the model. Results The average rodent density was 0.80% at the national surveillance sites in Shandong province in 2010-2018. The ARIMA model was used for the fitting of rodent density at the national surveillance sites in Shandong province in 2010-2018. With ARIMA (0,1,1)×(0,1,1)6 as the optimal model and white noise sequence as the residual sequence (χ2=0.035, P=0.832), this model was used to predict rodent density in each monitoring month of 2019, and the actual monitoring values fell within the 95% confidence interval of the predicted values. The seasonal variation trend of rodent density was basically consistent with the observed value, suggesting that the model could be used to predict rodent density in the short and medium term. Conclusion The ARIMA (0,1,1)×(0,1,1)6 model is well fitted with rodent density at the national surveillance sites in Shandong province and can be used for the prediction and early warning of rodent density, so as to provide a basis for the prevention and control of rodent-borne diseases.

Key words: Autoregressive integrated moving average model, Rodent density, Surveillance, Prediction

摘要: 目的 分析山东省国家级监测点(济南、青岛和聊城市)的鼠密度,探讨自回归移动平均(ARIMA)模型在鼠密度预测预警中的应用,方法 采用夹(笼)夜法监测各年鼠密度,应用R 3.6.2软件对2010-2018年各监测月份的鼠密度数据建立ARIMA模型,比较2019年各监测月份的预测鼠密度和实测鼠密度,评估预测效果。结果 2010-2018年山东省国家级监测点平均鼠密度为0.80%,应用ARIMA模型对山东省国家级监测点2010-2018年鼠密度进行拟合,选取ARIMA(0,1,1)×(0,1,1)6作为最佳模型,残差序列为白噪声序列(χ2=0.035,P=0.832),用此模型预测2019年的各监测月份鼠密度,实际监测值均落入预测值的95%置信区间内,且鼠密度季节消长趋势与实测值基本一致,可用于中短期内鼠类监测密度的预测。结论 ARIMA(0,1,1)×(0,1,1)6模型对山东省国家级监测点鼠类监测密度的拟合效果较好,可用于鼠密度的监测预警,为预防控制鼠传疾病提供依据。

关键词: 差分自回归移动平均模型, 鼠密度, 监测, 预测

CLC Number: