马尾松毛虫幼虫发生严重程度的预测研究
作者:
作者单位:

作者简介:

通讯作者:

基金项目:

国家林业公益性行业科研专项(201404410)资助。


Study on the forecasting occurence severity degree of the Dendrolimus punctatus larvae
Author:
Affiliation:

Fund Project:

  • 摘要
  • |
  • 图/表
  • |
  • 访问统计
  • |
  • 参考文献
  • |
  • 相似文献
  • |
  • 引证文献
  • |
  • 资源附件
    摘要:

    为了提高马尾松毛虫幼虫发生严重程度的预测精度,寻求简便准确的预测方法,采用时间平稳序列法、回归预测法、马尔科夫链法、BP神经网络法和列联表多因子多级相关分析法对安徽省潜山县1983—2014年的马尾松毛虫越冬代、一代和二代幼虫发生的严重程度进行预测,研究历史符合率,并用2015年和2016年的实际发生情况验证。结果表明,平稳时间序列法,列联表多因子多级相关分析法计算简便,预测结果准确;BP神经网络法和马尔科夫链法预测结果非常准确。回归模型中以当代卵盛期卵量预测当代幼虫发生严重程度的一元回归模型的预测结果准确性高,其余一元回归模型预测结果稍差,多元回归模型和逐步回归模型优于一元回归模型。BP神经网络模型是一种理想的预测模型。

    Abstract:

    In order to improve the prediction accuracy of the occurrence severity degree of the Dendrolimus punctatus larvae and find simple and accurate forecasting method. The method of stationary time series, regression?forecast, Markov chains, BP neural network and contingency table?analysis were applied to establish the prediction model of the occurrence severity degree of overwintering generation, the first generation and the second generation Dendrolimus punctatus larvae from 1983 to 2014 in Qianshan county of Anhui Province and it was used to study the historical coincidence rate, and then the predicted result was proved with the actual happening situaion in 2015 and 2016. Result shows that: The predicted results which used the more simple calculation methods of stationary time series and contingency table were accurrte; BP neural network and Markov chain method to predict the result were very accurate. The predicted result of single regression model that the contemporary amount of eggs from the egg stage predicted the occurence severity degree of the contemporary larvae was very accurate in the regression model and other predicted result with the single regression mode was a bit poor, so multiple regression model and the stepwise regression model were better than the single regression model. BP neural network model is a kind of ideal prediction model.

    参考文献
    相似文献
    引证文献
引用本文
分享
文章指标
  • 点击次数:
  • 下载次数:
历史
  • 收稿日期:
  • 最后修改日期:
  • 录用日期:
  • 在线发布日期: 2017-11-23