基于时间序列BP神经网络的集装箱吞吐量动态预测

集装箱吞吐量预测是港口发展规划制定的依据。在MATLAB环境下，把时间序列BP神经网络应用于港口集装箱吞吐量的预测，采用逐步递归的方法进行，同时注意尽量减少训练样本的浪费（只用1个检验样本）和充分挖掘BP神经网络适合短期预测的潜力。无论是从拟合情况，还是预测值的检验和港口发展规划的实际情况来看，都有着很高的精度．可以作为集装箱吞吐量预测的一种行之有效的方法。

Dynamic Prediction of Container Throughput Based on the Time Series BP Neural Network (BP NN)

Prediction of container throughput is the basis for making port development plan. The time series BP neural network was applied to the container throughput prediction of a port, by the method of step-by-step recursion, in the context of MATLAB. Meanwhile, minimizing the waste of the training samples and exploiting the potential of the short period prediction was paid more attention to. This method has a high precision in terms of either simulation, or the result of the prediction, thus is an effective method of prediction for port container throughput .