摘要
【目的】了解广西桑蚕原种微粒子病毒率(下简称毒率)的时间发展变化规律,为桑蚕微粒子病的防控提供理论依据。【方法】采用普通线性回归模型(Linear regression model)、自回归模型(Autoregression model)及季节性结构分量模型(Seasonal structure component)对2006~2010年广西桑蚕原种毒率的时间发展变化规律进行分析研究。【结果】普通线性回归模型对桑蚕原种毒率时间序列数据分析不理想;自回归模型能在一定程度上解决残差自相关问题,且预测值和实际值吻合程度较好;季节性结构分量模型分析结果表明,广西桑蚕原种病毒率在5年内(2006~2010年)呈现逐年上升的态势,且一年中桑蚕原种生产第1、4、5、6批的毒率较高,以第6批最高;第2、3、7、8、9、10、11、12批的毒率较低,其中以第8、9、10批次最低。【结论】时间序列分析能客观反映广西桑蚕原种毒率的发展变化规律,今后可从数据挖掘角度对桑蚕微粒子病进行防控。
[Objective]The present experiment was conducted to investigate the developmental dynamics of pebrine virus incidence on silkworm proto-species in Guangxi in order to effectively control silkworm pebrine disease.[Method]The time developmental dynamics of virus incidence on silkworm proto-species in Guangxi was worked out from 2006-2010 by adopting general linear regression model,auto-regression model and seasonal structure component model.[Result]It was found that general linear regression was not a satisfactory method to analyze time series data of virus incidences on silkworm proto-species.Auto-regression model,to some extent,resolved the self-relevant problems about residual error,and its predictive value conformed with actual value.The results obtained from seasonal structure component model revealed that virus incidence rates on silkworm proto-species in Guangxi were on the rising trend during the past five years(2006-2010) .In addition,the virus incidence rate of the 1s t,4th,5th and 6th batch of silkworm proto-species output was relatively higher with the highest in 6th batch,while that of the 2nd,3rd,7th,8th,9th,10th,11th and 12th batch was lower with the lowest in 8th,9th,10th batch.[Conclusion]Time series analysis can objectively reflect the development dynamics of virus incidence rate on silkworm proto-species in Guangxi.Silkworm pebrine disease can be prevented and controlled by using statistical prediction models in future.
出处
《南方农业学报》
CAS
CSCD
2011年第12期1560-1563,共4页
Journal of Southern Agriculture
基金
广西自然科学基金项目(2010GXNSFA013071)
关键词
桑蚕
原种
毒率
时间序列分析
自回归模型
季节性结构分量模型
silkworm
proto-species
virus incidence rate
time series analysis
auto-regressive model
seasonal structure component