基于灰色理论的高速公路小修保养维修量预测模型

在仅有通车5 a以下维修量的小样本数据条件下,为了合理、科学地预测高速公路小修保养维修量,量化年平均日交通量、年均降雨量、年均温度和通车年限4个因素对维修量的影响程度。提出了基于Matlab的灰色系统预测模型,以实际调研路段5 a的统计数据为样本,运用灰色关联度模型对4个因素的影响程度进行量化评估;分别建立均值GM(1,1)单变量预测模型与GM(1,N)多变量预测模型对维修量进行预测分析。研究结果表明:上述4个影响因素与维修量的灰色关联度均大于0.5,说明上述4个自变量对维修量均具有较大影响;GM(1,1)模型70%的预测结果误差小于15%,GM(1,N)模型75%的预测结果误差大于30%,从而得出GM(1,1)预测模型比GM(1,N)预测模型的预测结果更加准确。因此,对通车5 a以下维修量的小样本数据预测,选用单变量预测模型更加合理、有效。

Prediction Model of Minor Repair and Maintenance Amount of Expressway Based on Grey Theory

Under the condition of small sample data with less than 5 years'traffic volume,in order to predict the maintenance volume reasonably and scientifically,and quantify the influence degree of four factors on the maintenance volume,which are annual average daily traffic volume,annual average rainfall,annual average temperature and service life,this paper puts forward the grey system Expressway Based on MATLAB.Taking the 5-year statistic data of the actual investigation sections as the sample,the grey relational degree model is used to quantify the influence degree of four factors on the maintenance volume,namely,the average daily traffic volume,the average annual rainfall,the average annual temperature and the length of service,and the GM(1,1)univariate prediction model is established.And the GM(1,N)multivariate prediction model is used to predict the maintenance volume.The results show that the grey correlation degree of the above four factors is greater than 0.5,indicating that the above four independent variables have a greater impact on the maintenance volume;under the condition of small sample data of less than 5 years of traffic,the error of GM(1,1)univariate prediction model is mostly less than 15%,Most of the errors of the predicted results of the GM(1,N)multivariate prediction model are more than 30%.Thus,the predicted results of the GM(1,1)univariate prediction model are more accurate than those of the GM(1,N)multivariate prediction model.Therefore,it is more reasonable and effective to select the univariate prediction model for the prediction of the road maintenance volume under 5 years of traffic.