为能更加精准预测不同时段的航班延误态势，选用美国亚特兰大机场2023年全年的运行数据与相关气象数据进行实验，提出基于相关系分析（correlation analysis, CA），主成分分析（principal components analysis, PCA）和Informer模型的CA-PCA-Informer航班延误预测模型，采用MAE（mean absolute error）和RMSE（root mean square error）作为模型的评价指标进行预测误差分析。结果表明，CA-PCA-Informer模型比简单的组合模型预测效果更好，与CA-PCA-LSTM和CA-PCA-GRU模型相比模型误差最低，MAE和RMSE分别降低了20.2% ~ 20.7%和12.7% ~ 14.1%；CA-PCA-Informer模型对预测步长为1 h时预测更为精准，该模型可以为决策者提供更加准确的航班延误态势以保证航班的高效运行。

In order to more accurately predict flight delays at different times of the year, flight delay prediction trends was investigated using operational and meteorological data from Atlanta Airport in the United States for the year 2023. A CA-PCA-Informer flight delay prediction model, incorporating correlation analysis (CA), principal component analysis (PCA), and the Informer model, was proposed. Mean absolute error (MAE) and root mean square error (RMSE) were utilized as evaluation metrics to assess the prediction error. The findings reveal that the CA-PCA-Informer model outperforms simpler combined models, demonstrating the lowest error compared to the CA-PCA-LSTM and CA-PCA-GRU models, with MAE and RMSE reductions of 20.2% ~ 20.7% and 12.7% ~ 14.1%, respectively. The CA-PCA-Informer model is particularly effective for one-hour ahead predictions, providing decision-makers with more accurate flight delay trends to enhance efficient flight operations.