论文部分内容阅读
为了充分利用航空运输事故征候的数据信息、提高预测精度,采用最优变权组合模型对其进行预测。以灰色Verhulst模型、Brown指数平滑模型及非线性回归模型为单项模型,以样本点处组合预测误差绝对值最小为原则,构建航空运输事故征候的最优变权组合预测模型。根据2002—2011年的相关数据,分别利用单项模型和组合模型预测事故征候万时率。结果表明:未来2 a事故征候万时率分别为0.347和0.331,呈下降趋势;最优变权组合模型能够克服灰色Verhulst模型和回归预测模型不能反映数据波动性及指数平滑预测滞后性的缺陷,其预测稳定性和精度都高于单项预测模型。
In order to make full use of the data and information of air transport accidents and improve the prediction accuracy, the optimal weight combination model is used to predict the data. Taking the gray Verhulst model, the Brown exponential smoothing model and the nonlinear regression model as the single models, the optimal combination weighting forecasting model of air transport incident is constructed based on the principle of minimizing the absolute value of the combined forecasting error at the sample points. According to the relevant data from 2002 to 2011, the single-item model and combined model are respectively used to predict the accident rate. The results show that in the next 2 years, the hourly rates of incidents are 0.347 and 0.331, respectively, which show a decreasing trend. The optimal combination of weights can overcome the shortcomings that the gray Verhulst model and the regression prediction model can not reflect the data volatility and exponential smoothing prediction lag. Its prediction stability and accuracy are higher than the single prediction model.