达乌尔黄鼠鼠疫预报研究

发布时间：2018-05-03 06:31

本文选题：达乌尔黄鼠 + 鼠疫　；参考：《第三军医大学》2006年硕士论文

【摘要】： 目的: 通过对影响达乌尔黄鼠鼠疫流行因素的分析,建立达乌尔黄鼠鼠疫预报模型,为制定鼠疫预防和控制措施提供依据。方法: 1.相关分析。收集吉林西北部草原1957～2005年4～8月份的鼠疫监测数据及气象资料,利用spss13.0 for windows软件对资料及鼠疫流行强度进行相关分析,选取影响鼠疫流行的相关因子进行模型的建立。 2.建立神经网络模型。运用Matlab7软件包中的神经网络工具箱,根据吉林西北部草原1957～2005年达乌尔黄鼠密度、黄鼠体外蚤指数、黄鼠体外染蚤率、黄鼠体外蜱指数、月平均气温、月平均相对湿度、月平均气压、月平均风速、月平均降水量、灭鼠后黄鼠密度、小型鼠捕获率等数据建立达乌尔黄鼠鼠疫流行的自组织映射网络预测模型,并对模型进行验证。结果: 1.相关分析结果。相关分析结果显示,达乌尔黄鼠密度、黄鼠体蚤指数、黄鼠洞干蚤指数、黄鼠体蜱指数、平均气温、平均相对湿度、平均风速、降水量、灭鼠后黄鼠密度、小型鼠捕获率与黄鼠鼠疫流行强度呈正相关,平均气压与黄鼠鼠疫流行强度呈负相关。 2.神经网络模型。SOM网络经学习和训练后,得出聚类分级结果,输出结果与鼠疫流行强度的真实值符合率达93.3%,通过建立的神经网络模型对未来鼠疫流行强度进行预测,预测结果与实际结果相符。结论: 1.自组织映射的神经网络对鼠疫流行的预测效果较好,具有较高的应用和推广价值,可作为鼠疫流行预测的一种新方法。 2.鼠疫流行影响因子较多,要建立更加准确的预测预报模型,需要更全面、准确收集数据。
[Abstract]:Objective: Based on the analysis of the factors affecting the epidemic of the plague of the Daour yellow rat, the forecasting model of Daour plague was established, which provided the basis for the prevention and control measures of the plague. Methods: 1. Correlation analysis. The monitoring data and meteorological data of plague in northwest grassland of Jilin Province from 1957 to 2005 were collected. The data and the intensity of plague epidemic were analyzed by spss13.0 for windows software, and the relevant factors affecting plague epidemic were selected to establish the model. 2. The neural network model is established. Using neural network toolbox of Matlab7 software package, according to the density of Daour yellow rat, the index of flea in vitro, the rate of flea infection in vitro, the index of tick in vitro, the monthly mean air temperature and the average relative humidity of Daour in northwest grassland of Jilin Province from 1957 to 2005. Based on the data of monthly mean air pressure, monthly average wind speed, monthly average precipitation, rodent densities after rodent control and capture rate of small mice, a self-organizing mapping network for predicting the prevalence of Daour yellow rat plague was established, and the model was verified. Results: 1. Correlation analysis results. The results of correlation analysis showed that the density of Daour, flea index, dry flea index, tick index, mean air temperature, mean relative humidity, mean wind speed, precipitation, density after rodent control, The capture rate of small mice was positively correlated with the epidemic intensity of Yersinia pestis, and the mean air pressure was negatively correlated with the epidemic intensity of Yersinia pestis. 2. After learning and training, the neural network model, SOM network, obtained the clustering classification result, the output result and the plague epidemic intensity true value coincidence rate reached 93.30.The future plague epidemic intensity was forecasted by the established neural network model. The predicted results are in agreement with the actual results. Conclusion: 1. The self-organizing mapping neural network has a good effect on the prediction of plague epidemic and has a high value of application and popularization. It can be used as a new method for plague epidemic prediction. 2. There are many factors affecting plague epidemic, so it is necessary to collect more comprehensive and accurate data in order to establish more accurate forecasting model.
【学位授予单位】：第三军医大学
【学位级别】：硕士
【学位授予年份】：2006
【分类号】：R181.3

【相似文献】