摘要
路面性能预测方法丰富多样,各有其局限性和适用性。对现有的几个典型的预测方法和模型进行对比研究,以建立最优的高速公路路面性能预测模型。针对高速公路路面历史数据资料较少以及路面性能衰减趋势的随机性、影响因素的不确定性,采用移动平均法、指数平滑法、灰色系统理论、灰色马尔可夫模型,分别建立路面性能预测模型进行对比,结合高速公路实际检测数据进行了模型预测计算和验证,并对预测结果进行了初步分析。该实验结果得出时间序列法以及灰色模型预测效果较好,其中时间序列移动平均法的预测相对误差在15%范围以内。
Pavement performance prediction methods are diversified, each of which has its own limitations and applicability. Several existing typical prediction methods and models were compared and researched to establish an optimal highway pavement performance prediction model. By focusing on insufficient historical data on highway pavement, randomness of the tendency of pavement performance attenuation and uncertainty of influencing factors, moving average method, exponential smoothing method, grey system theory and grey-Markov model were used to separately establish pavement performance prediction models for comparison. On actual highway test data, model prediction computation and verification were carried out and prediction outcomes were preliminarily analyzed. The experimental results conclude that time series method and grey model have relatively good prediction effects, in which, the relative error of prediction by using time series moving average method is less than 15%.
作者
樊文有
许玉凤
陈菡
戴龙
程婷
FAN Wenyou;XU Yufeng;CHEN Han;DAI Long;CHENG Ting(School oflnformation Engineering,China University of Geosciences,Wuhan Hubei 430074,China;The Greatmap Company Limited,Beijing 100085,China)
出处
《计算机应用》
CSCD
北大核心
2018年第A01期9-12,27,共5页
journal of Computer Applications
基金
国家863计划项目(2014AA123001)
