Fuel consumption prediction for pre-departure flights using attention-based multi-modal fusion

Yi Lin, Dongyue Guo, Yuankai Wu, Lishuai Li, Edmond Qi Wu, Wenyi Ge

Abstract: Highlights•The FCPNet is proposed to support the fuel loading decisions before flight departure.•The GCN- and ConvLSTM-based networks are used to consider multi-modal inputs.•Attention-based fusion is to learn task-oriented features for the fuel consumption.•A binary encoding representation is proposed to enhance the training convergence.•The proposed model can achieve comparable performance over post-prediction methods.