摘要
在对现有各种预测方法进行简要回顾的基础上 ,对遗传规划方法进行了研究 ,提出了应用于集装箱吞吐量预测的计算模型和相关参数的确定方法 ,并完成了算法设计和编程工作 .在此基础上 ,以宁波港集装箱吞吐量的历史数据为样本 ,开展了预测计算试验 ,找到了能够反映宁波港目前集装箱吞吐量发展规律的样本集 ,并对宁波港未来九年的集装箱吞吐量进行了预报 .结果表明 ,本算法合理、改良自动化程度高 ,为解决集装箱吞吐量等非线性系统预测提供了一条新的途径 .
This paper focused on more scientific and reasonable prediction method of containers handling capacity. Various prediction methods were reviewed and the genetic programming was briefly introduced. The different relative factors such as function sets, terminator sets, possibilities of reproduction, crossover and mutation, termination precision were studied and tested, and the final model which was applied in the prediction of containers handling capacity was established and the algorithm was programmed, Taking the containers handling capacity of Ningbo Port as an example, the different example data from historical data were selected for computation tests, and the sample data, which was used in the prediction of the future containers handling capacity of Ningbo Port, was determined. Then, the handling capacities of the coming nine years were predicted. The application result shows that the prediction method is reasonable and easy to achieve self-improvement. It is a new way to resolve the prediction of nonlinear system such as prediction of containers handling capacity.
出处
《上海交通大学学报》
EI
CAS
CSCD
北大核心
2003年第8期1246-1250,共5页
Journal of Shanghai Jiaotong University
关键词
遗传规划
预测方法
非线性映射
集装箱吞吐量预测
Distribution of goods
Flowcharting
Genetic algorithms
Mapping
Nonlinear systems