【摘要】 随着中国民航业的高速发展，运输航空器的碳排放问题逐渐引起关注。采用时间序列的方法建立了传统的差分整合移动平均自回归（ARIMA）模型以及优化的长短期记忆网络（LSTM）模型，对航空器碳排放量、碳排放强度以及吨公里碳排放量进行了预测，通过鲸鱼优化算法（WOA）对LSTM中的学习率和隐藏节点数进行优化，避免了人为选择参数的主观性和盲目性，有利于提高模型预测的准确性。通过对比两种模型的均方根误差（RMSE）和平均绝对误差（MAE）,ARIMA模型在航空器碳排放预测中有较好表现，WOA-LSTM模型在碳排放强度、吨公里碳排放的预测中有较好表现。更多还原

【Abstract】 With the rapid development of China’s civil aviation industry, the carbon emission of transport aircraft has gradually attracted attention. The research uses the time series method to establish the traditional Autoregressive Integrated Moving Average(ARIMA) model and the optimized Long Short Term Memory(LSTM) model to predict the carbon emissions, carbon intensity and carbon emissions per ton kilometer of the aircraft. The learning rate and hidden node number in LSTM are optimized through the Whale Optimization Algorithm(WOA), avoiding the subjectivity and blindness of manually selecting parameters, it is helpful to improve the accuracy of model prediction. By comparing the Root Mean Squared Error(RMSE) and Mean Absolute Error(MAE) of the two models, ARIMA model performs well in aircraft carbon emission prediction, and WOA-LSTM model performs well in carbon intensity and ton kilometer carbon emission prediction.更多还原