邱颖;陆化普;王海威;

• 1:Institute of Transportation Engineering
• 2:Department of Civil Engineering
• 3:Tsinghua University

摘要(Abstract)：

Currently applied prediction methods of regional freight traffic and freight ton-kilometer forecasting were analyzed using typical Chinese regional goods transportation characteristics. The review of prediction methods showes that practical planning experts tend to apply the traditional methods which are easier to implement. The comparison also demonstrates that a combination of traditional methods is more effective than the simple models for practical planning. Research using the statistical data for the Yangtze Delta, Pearl River Delta, and Bohai Rim areas shows that ignoring differences between transport modes impacts the pre-diction accuracy. The four main transport modes suit different methods. The results show that the power model is better for railways, and the linear model is better for highways and waterways. Thus a combined model gives better results for all modes. The results for regional systems can be generalized to national transportation systems.

关键词(KeyWords)：

Abstract:

Keywords:

基金项目(Foundation): the National High-Tech Research and Development (863) Program of China (No. 2007AA11Z202)

作者(Authors): 邱颖;陆化普;王海威;

参考文献(References)：

• [1] Chen Yuxiang, Zhang Hanya. Prediction Technology and Application. Beijing: Mechanical Industry Press, 1985. (in Chinese)
• [2] Pandit S M, Wu Xianming. Time Series and System Analysis with Application. New York: Wiley, 1983.
• [3] Carbon M, Delecroix M. Non-parametric vs parametric forecasting in time series: A computational point of view. Applied Stochastic Models and Data Analysis, 1993, 9(3): 215-229.
• [4] An Hongzhi, Gu Lan. Fast stepwise procedures of selection of variables by using AIC and BIC criteria. Acta Mathe- maticae Applicatae Sinica, 1989, (5): 60-67.
• [5] Gong Lisha. Semiparmetric regression model wavelet es- timation [Dissertation]. Changsha: Hunan University, 2003. (in Chinese)
• [6] Wen Kunli. Grey systems: Modeling and prediction. USA: Yang’s Scientific Research Institute, 2004.
• [7] Jiang Bo. Research on analytical method with grey system and neural network and their application [Dissertation]. Wuhan: Huazhong University of Science and Technology, 2004. (in Chinese)
• [8] Rovithakis G A, Christodoulou M A. Adaptive control of unknown plants using dynamical neural networks. IEEE Trans. Syst. Man Cyber, 1994, 24(3): 400-412.
• [9] Huang Qingxu. Self-adapt arithmetic coding method based on the neural network and maximum entropy [Dissertation]. Chengdu: Chengdu Technology University, 2004. (in Chinese)
• [10] Stolarski T A. A probabilistic approach to wear prediction. Journal of Physics D: Applied Physics, 1990, 23: 1143-1149.
• [11] Weng Xiaoxiang, Luo Jian. Using fractal method in the filed of economic forecasting. Statistics and Information Forum, 2005, 20(5): 39-42. (in Chinese)
• [12] Lian Dajun, Yuan Ming. The dynamic data forecasting method and its application based on power spectrum esti- mation by maximum entropy method. Journal of Univer- sity of Science and Technology of Suzhou (Engineering and Technology), 2005, 18(3): 38-40. (in Chinese)
• [13] Falconer K J. Fractal Geometry: Mathematical Founda- tions and Application. New York: Wiley, 1990.