Predicating the Coastal Throughput Based on Gray RBF Neural Network
Usually, if one uses the traditional econometric models to predicate the coastal throughput, its always difficult to get satisfactory results. On one hand, there are many factors affecting forecast, on the other hand the data are abnormal and mercurial owing to the economic environment and policy implications. However, RBF neural network can solve such problems, which have the ability to approximate arbitrary nonlinear mapping through learning. This paper establishes a RBF neural network model, which puts the gray time series as the networks input layer. At last, it takes the cargo throughput in Fujian Province for example to verify the model. The results show that it can realize throughput predication much better and the projections has higher accuracy.
coastal throughput gray time series RBF neural network
Zhang Wenfen Liu Qing
School of Transport, Wuhan University of Technology Wuhan, China
国际会议
成都
英文
1845-1848
2010-12-17（万方平台首次上网日期，不代表论文的发表时间）