中文摘要：

针对终端空域内短期天气条件变化引起的航班延误问题，首先利用雷达回波平面位置显示（PPI）和距离高度显示（RHI）对降水等级分类．通过图像识别方法对不同降水等级的云团进行轮廓提取；其次将对应时间的交通流信息与降水云团的空间范围匹配．提出了不同降水等级条件下的三维天气影响交通指数（WITI）计算方法：最后结合实际管制规则，将WITI指数与管制间隔调整和改航条件下的延误时间作指数回归分析，得到短期延误预测模型。通过算例仿真验证了该模型的拟合优度相比于线性回归模型提高了1．6％～4.3％，且估计标准误差趋近于0,说明指数回归延误预测模型更能反映天气影响下终端空域的实际延误状况。

英文摘要：

Aiming at the flight delay in terminal airspace caused by the changes of short-term weather conditions, radar e- cho PPI （plane-position indication ） and RHI （range-height indication ） are firstly used to classify the precipita- tion grades, then the contour of different grades of convective clouds is extracted by image recognition method. Secondly, the corresponding time of traffic flow information is matched to the convective clouds, and the calcula- tion method of 3D WITI （weather impacted traffic index） based on different levels of precipitation is put forward. Finally, according to air traffic control rules, a short-term delay prediction model is proposed through exponen- tial regression analysis between WITI and delay time under different control intervals or rerouting path distance. Simulation resuhs show that under the condition of control interval adjustment and rerouting, compared with the linear regression model, the fit goodness of exponential regression model increases 1.6%-4.3%, and the standard error of estimate is closed to 0, indicating that the exponential regression delay prediction model can better re- flect the actual delay situation of terminal airspace under the influence of precipitation.