Research on a Comprehensive Prediction Model to Determine the Number of Expressway Lanes

Recently, China’s transportation capacity has posed great challenges: whether the prediction for the number of lanes in the design stage is in line with the fact to study has brought about public attention. Therefore, we first analyzed China’s existing modeling method and afterward proposed a comprehensive model of “grey degree-principal components-neural network” model to determine the number of expressway lanes, which is referred to as “three-step prediction model.” Finally, we verified it using relevant data of Changchun-Jilin Expressway in Jilin Province, China. The model comprehensively took into account various relevant factors influencing traffic volume generation and compensated the shortcoming of simple auto-regressive analysis based on historical statistics of road traffic volume used in the traditional methods such as the growth rate method, exponential smoothing method, and grey degree method. Overall, “three-step prediction model” is a great significance to avoid duplicate construction and over-construction ahead of time in highway construction.