成都市蚊密度中的混沌现象

The chaos phenomenon in mosquito density of Chengdu City|

目的2006—2021年成都市城区蚊密度呈现宽幅震荡趋势,可将其一维时间序列拓展到多维相空间去进行混沌分析,以确定数据序列的非线性特征。方法使用L.Cao算法确定嵌入维数和关联维数,后对数据序列进行相空间重构,由雅可比(Jacobian)方法计算最大李雅普诺夫(Lyapunov)指数。结果当自相关函数衰减为0时,成都市蚊密度数据序列嵌入时延为τ=2,estimateEmbeddingDim函数确定的嵌入维数为m=7,系统的关联维数介于2~3。lyapunov.max函数计算所得的最大李雅普诺夫指数为λ_(1)=0.024,证明该动力系统至少有1个李雅普诺夫指数大于0,提示蚊密度有混沌特征。结论成都市蚊密度分析可使用非线性方法,预测结果能较好地表征城区蚊密度的变化趋势和规律,对20个月内蚊密度中混沌现象短期预测精度较高。

Objective A broad oscillation trend was shown for mosquito density in the urban area of Chengdu City from 2006 to 2021,and its one-dimensional time series could be extended to multi-dimensional phase space for chaos analysis to determine the nonlinear characteristics of the data series.Methods The L.Cao algorithm was adopted to calculate the embedding dimension and correlation dimension,and then the phase space reconstruction of the data sequence was performed.The Jacobian method was used to calculate the maximum Lyapunov exponent.Results When the autocorrelation function decayed to O,the embedded delay of the mosquito density data series in Chengdu City wasτ=2.The embedding dimension determined by the estimate Embedding Dim function was m=7,the Correlation dimension of the dynamical system was between 2-3.The maximum Lyapunov exponent calculated by the lyapunov.max function wasλ_(1)=0.024,proved that the dynamical system had at least one Lyapunov exponent greater than O indicating that the mosquito density had chaotic phenomena.Conclusion The nonlinear method can be used to analyze mosquito density in Chengdu City,and the prediction results can better characterize the trend and patern of mosquito density in the urban area,and the short-term prediction accuracy of the chaotic phenomenon in mosquito density in 20 months is high.