【摘要】 [目的]检测马尾松蛀干类害虫2006～2010年,林间种群数量是否具有混沌特性。[方法]利用前馈神经网络分析马尾松蛀干类害虫林间种群数量的复杂性动态。[结果]前馈网络模型估计的最大Lyapunov指数为0.0128,说明马尾松蛀干类害虫林间种群序列存在混沌现象。[结论]马尾松蛀干类害虫的数量与前一次或前几次观测值密切相关,可用重构相空间的方法预测下一次观测值。更多还原

【Abstract】 [Objective] To detect the chaos characteristic of population quantity of Pinus massoniana trunk borers from 2006 to 2010.[Method] The feedforward neutral network approach was adopted to analyze the complex dynamics of P.massoniana trunk borers.[Result] The largest Lyapunov exponent estimated by feedforward network model was 0.012 8,indicating the chaos features of the population sequence of P.massoniana trunk borers.[Conclusion] The present quantity was closely correlated with the previous observation values,P.massoniana trunk borers could be forecasted by reconstructed phase space method.更多还原