高速铁路车站岔区高路堤沉降组合预测研究

有砟轨道高速铁路车站岔区由于车轨系统动力相互作用复杂，线路几何形位变化较快，且维修难度大。为提高线路稳定性，可将车站岔区有砟轨道更换为无砟轨道。依托中兰客专北滩车站岔区铺设无砟轨道的工程，分别对路基补强方案和路基沉降预测开展研究。为满足岔区铺设无砟轨道的要求，提出路基冲击碾压、堆载预压及路基面封闭隔水的补强方案。基于灰色Verhulst、双曲线和邓英尔预测模型的共性和个性，对三者进行最优化组合。以组合模型的最小误差平方和为目标函数求解最优加权系数，构建组合预测模型。通过预测对比发现，三模型组合预测模型的精度优于任一单一模型和两模型组合模型，且适应性更强更可靠。

Study on Combined Predication of High Embankment Settlement in Turnout Junction Area of High Speed Railway Station

At the turnout junction area of high-speed railway of ballast track, due to complex dynamic interaction of the track system, the track geometry changes so rapidly that the maintenance thereof is difficult. In order to improve the stability of the line, the ballast track in the turnout junction area may be replaced by ballastless track. Here based on the engineering of ballastless track laying in the turnout junction area at Beitan Station of Zhongwei—Lanzhou Passenger Dedicated Line, the subgrade reinforcement scheme and subgrade settlement prediction were studied respectively. In order to meet the requirements of ballastless track laying in turnout junction area, the reinforcement scheme including impact rolling and surcharge preloading of subgrade, and closed subgrade surface for water resistance is proposed. Based on the commonness and individuality of grey Verhulst, hyperbola and Deng Yinger prediction models, the optimized combination is carried out. The sum of minimum squared error of the combined model is taken as the objective function to solve the optimal weighting coefficient, and then to construct the combined prediction model. Through the comparison of prediction, it is found that the accuracy of the prediction model combined by the three models is better than that of any single model and any model combined by two models, moreover, it has stronger and more reliable adaptability.