基于ELM模型的民航客运量预测研究

Research on Passenger Traffic Forecasting Based on ELM Model

航空企业在做规划运营之前,首要是要对民航客运量进行精准预测,这是航空公司在做重大科学决策和实施可行性计划的重要保障。然而,影响民航客运量的因素复杂多样,传统的预测方法难以达到日益高要求的预测精度,据此,文章基于影响民航客运量主要因素,一年的国内生产总值、外国人入境游客、定期航班航线里程、铁路客运量、第三产业增加值,利用超极限学习机(Extreme Learning Machine,ELM)的算法模型,对民航客运量进行预测,结果表明基于ELM预测模型具有较好的预测精度。

Before the aviation enterprises do planning and operation,the first priority is to accurately forecast the passenger traffic volume of the civil aviation.This is an important guarantee for the airlines to make major scientific decisions and implement feasible plans.However,the factors affecting passenger traffic in civil aviation are complex and diverse,and traditional prediction methods are difficult to achieve increasingly high accuracy prediction accuracy.Accordingly,based on the main factors affecting passenger traffic in civil aviation,one year's gross domestic product,foreign inbound tourists,regular flight route mileage,railway passenger traffic,and tertiary industry added value,based on the algorithm model of extreme learning machine(ELM),the passenger traffic volume is predicted.The results show that the ELM prediction model has better prediction accuracy.