本文旨在研究通货膨胀预期与实际通货膨胀之间的影响关系，在数据获取及预处理阶段，将从国家统计局取得的居民消费物价指数CPI加工成为实际通货膨胀率，再将从中国人民银行《全国城镇居民储蓄问卷调查综述》取得的数据处理成为预期通货膨胀率，这一处理方法采用国际通用的差额法和概率法。在通货膨胀率预测阶段，首先采用单一的预期数据预测，效果不太理想，然后采用传统时间序列ARMA模型对单一的实际数据预测，效果有所改善，仍不太理想，最后将二者结合，使用自回归分布滞后ADL模型对预测进行改进，结果显示，预测精度有了大的提高。因此，本文的主要结论是：通胀预期对实际通货膨胀率的形成无决定性影响，但有较为明显的影响。故而一国应当加强通货膨胀预期管理研究，在通货膨胀来临之前进行提前管理和引导。当然，本文的模型并非最佳，很多地方显示该法的不足，要做到更好的预测，还需考虑更多的因素。

This thesis studies the influence relation between the inflation expectation and the inflation rate.In data acquisition and preprocessing stage,we get real inflation out of the monthly CPI from National Bureau of Statistics,and get the inflation expectation out of the Consumers' Saving Survey which is conducted by People's Bank of China.In the inflation forecasting phase,using only the expectation data or the real data is unsatisfactory.Therefore,we combine these two kinds data to build an ADL model,the results shows that the prediction accuracy has been improved.The main conclusion is that the inflation expectation has no decisive,but obvious effects to the real inflation rate.It reminds the government to pay attention to the management of inflation expectation. Of course, this model is not optimal, to achieve better prediction, we also need to consider more factors.