论文部分内容阅读
受实测数据变异性大的影响,路面实测温度的丰富信息往往未得到充分挖掘,现有的路面温度预测模型也大多将变异性当作“纯随机性”而不加考虑。为充分揭示路面实测温度的分布规律,剖析实测温度的确定性和随机性成分,构建随机性路面温度预估模型,利用某高速公路以1h为间隔的实测温度数据,一方面在时域层次分析其基本的统计学特性,另一方面在频域层次分析其频谱构成。在此基础上,采用随机论的相关方法,建立了基于ARIMA乘积结构的路面温度预估模型。结果表明,路面温度的随机性成分不可忽略,所建模型能够顺利通过残差的白噪声检验,ARIMA乘积结构适用于路面的温度预测。
Due to the large variability of the measured data, the abundant information of the measured temperature of the pavement is often not fully tapped, and the existing pavement temperature predictive model mostly regards the variability as “purely random” without considering it. In order to fully reveal the distribution law of the measured temperature in the pavement, the deterministic and stochastic components of the measured temperature are analyzed. A stochastic prediction model of the pavement temperature is established. Based on the measured temperature data of an expressway at intervals of 1h, Its basic statistical properties, on the other hand, analyzes its spectrum composition in the frequency domain. On this basis, using the stochastic theory method, a prediction model of pavement temperature based on ARIMA product structure is established. The results show that the random components of pavement temperature can not be neglected. The model can successfully pass the residual white noise test. The ARIMA product structure is suitable for the temperature prediction of pavement.