改进的RBF神经网络在港口集装箱吞吐量预测中的应用

Application of Improved RBF Neural Network for Port Container Throughput Forecast

本文首先分析了RBF神经网络结构的基础,指出基函数中心的确定是影响网络精度的重要因素。在分析常用中心确定算法的基础上,提出了一种改进算法,给出了算法的具体实现过程。并将改进的RBF算法用于对港口集装箱吞吐量的预测评价,应用成果表明此优化神经网络用于集装箱吞吐量预测结果更加合理,精度有很大提高,有着广泛的应用前景。

This paper analyzes the structure of RBF neural network first, and points out that determination of basis function center is the key factor affecting the precision of network. Based on the analysis of common algorithm for the center, this paper presents an improved algorithm, together with its specific realization process. The result from the actual application in forecasting the container throughput shows that it is more effective and has a broad application prospect.