航班的配餐量预测方法、装置、计算机设备和存储介质
By using time series analysis and attention mechanisms, flight time series are generated, the residual of the number of passengers booked is encoded, weights are calculated, and the prediction model is optimized. This solves the problem of inaccurate prediction of the number of passengers departing from flights and achieves efficient and accurate prediction of meal quantities.
CN115438835BActive Publication Date: 2026-07-17XIAMEN AIRLINES CO LTD
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- XIAMEN AIRLINES CO LTD
- Filing Date
- 2022-08-03
- Publication Date
- 2026-07-17
AI Technical Summary
Technical Problem
Current technology cannot accurately predict the number of passengers departing from a flight, resulting in inaccurate predictions of in-flight catering needs, which affects the airline's service quality and cost control.
Method used
Using time series analysis and attention mechanisms, the system generates flight time series, counts the number of reservations for each time period, encodes the residuals of the number of reservations, calculates the weights of the number of reservations, and uses an optimized meal preparation prediction model for prediction.
Benefits of technology
It improved the accuracy and efficiency of flight catering quantity prediction, solved problems such as gradient explosion, and achieved efficient and accurate catering quantity prediction.
✦ Generated by Eureka AI based on patent content.
Smart Images
Figure CN115438835B_ABST
Abstract
本申请涉及一种航班的配餐量预测方法、装置、计算机设备、存储介质和计算机程序产品。所述方法包括：根据时间序列长度生成航班的时间序列，统计时间序列中的各时间段的订座人数；分别对各时间段进行编码，并通过依次编码各时间段的订座人数的残差生成各时间段的隐藏层向量；将各时间段的编码和各时间段的隐藏层向量进行注意力机制计算，得到各时间段的订座人数权重；对各时间段的订座人数权重进行解码，基于解码的结果确定配餐量预测值；基于配餐量预测值和配餐量标签值之间的误差对配餐量预测模型进行优化，利用优化后的配餐量预测模型对航班的配餐量进行预测。采用本方法能够通过注意力机制和残差的编码，提高配餐量预测的准确性。
Need to check novelty before this filing date? Find Prior Art