高速公路小修保养工程量的预测模型

Expressway minor maintenance amount prediction based on neural network

为了科学合理地预测高速公路小修保养工程数量,分析了影响高速公路小修保养工程量的主要因素,提出了基于Matlab的BP神经网络高速公路小修保养工程量预测方法;以沥青路面年小修保养工程量历史统计数据为样本,将路面质量等级评分值、路面使用性能指数、年均日交通量、年均日重车量、计算使用年限、年均降雨量、年平均温差、路面厚度、路面宽度确定为影响沥青路面小修保养工程量的量化指标,建立了预测沥青路面小修保养工程量的多元非线性模型;以调研路段沥青路面实际破损维修数量为样本,分别用BP神经网络模型和多元线性回归模型进行预测分析。研究结果表明:将预测结果与实际维修量数据进行比较,BP神经网络模型预测误差为5%,多元线性回归模型预测误差为14%,说明BP神经网络模型是预测高速公路沥青路面小修保养工程量的一种较为合理可行的方法。

In order to predict expressway minor repair and maintenance amount scientifically and reasonably, the main factors of expxessway minor repair and maintenance amount were analyzed and the maintenance amount prediction method of BP neural network based on Matlab was put forward. With the samples of historical statistics of asphalt pavement minor maintenance a- mount, the quantitative indexes of factors including the score value of pavement quality, the pavement performance index, the annual average volume of daily traffic, the annual average of daily heavy vehicles, the road service life, the annual mean temperature difference, the pavement thickness and the pavement width were under consideration. Prediction model of pavement minor repair and maintenance was established. According to pavement distress amount of investigated expressways, analyses and predictions were made respectively by the BP neural network model and multiple linear regression model. The results show that compared with the actual data, the prediction error of BP neural network model is 5 %, and the prediction error of multiple linear re- gression model is 14%, which indicates that BP neural network model to predict the minor repair and maintenance amount of expressway asphalt pavement is a more reasonable and feasible meth- od. 3 tabs, 1 fig, 11 refs.