时间序列预测的发展与应用

Development and Application of Time Series Forecasting

为了给运载火箭测试数据预测问题的研究打下基础,开辟研究思路,对时间序列预测方法进行研究。介绍了传统时间序列预测方法中的指数平滑法、季节模型和求和回归滑动平均模型的发展,重点阐述了时间序列预测的3类热门研究领域:多变量时间序列模型、模糊时间序列模型和组合预测模型,并通过文献介绍,分析了时间序列预测技术的应用与改进方向。

In order to lay a foundation for the data forecasting of the rocket launch test and open up some ideas, the time series forecast methods are researched. The development of exponential smooth, seasonal method and autoregressive integrated moving average model, which are part of the traditional time series forecasting method, are introduced at first. And then three hot areas of time series forecasting are introduced, which is multivariate time series model, fuzzy time series mode and combination forecasting model. The future direction of application and improvement of time series forecasting are analyzed through the introduction of literatures.