Determining the reasonable scale of highway mileage is an important step for the realization of highway network planning, and it is the premise and foundation of network optimization. To address the problem of reasonable scale of China's highway mileage prediction method of variety and unification, this study aims to improve the accuracy of the reasonable scale of highway mileage of the predicted value and enhance the rationality and reliability by analyzing the factors affecting the development of highway mileage systematically to determine the regional population, area, and economic development of the state, car ownership, and the relating level of development of highway mileage. Then we used factors of population, area, car ownership, and other economic development to correct the highway mileage growth curve forecasting model and obtain the social and economic growth curve forecasting model. Finally, we used historical data from 2000-2012 in the Shandong Province to verify the accuracy of the constructed model. The results show that the error of the socio-economic growth curve forecasting model can be controlled within 4%. Finally, we used the model to predict the reasonable scale of highway mileage of the Shandong Province.

This paper introduces a new model for a single-unit non-Markov repairable system in which the repair time is sufficiently short (less than some given critical value); it won't result in system failures. Such short repair interval can be omitted from the downtime record. In this paper the critical repair time is considered into two classes, one is a constant and the other is a non-negative random variable. Then the availability for the new model is calculated. Finally, some numerical examples are given to illustrate the proposed model.