A WITI Based Airport Delay Estimation Using Piece-wise Linear Model

With the increasing of air traffic demand,the improvement of efficiency of air transportation operation have become a very important issue.Recently,the researches indicate that the impact of severe weather on air traffic is the biggest reason which leads to air traffic delay,so the study based on impacts caused by severe weather on air traffic can benefit the efficiency of air traffic as well as provide a good solution and pre-coping mechanisms for airport delay.The paper established a delay prediction model which is mainly focus on the airport and its surrounding airspace.The concept weather Impact air traffic index(WITI)is used.To compatible to this research environment,wind speed,visibility and significant weather are chosen as the parameters to model WITI,the coefficients of these parameters are measured using regression model.A prediction model is generated by fit WITI and real delay data in a regression equation,Considering the variation relation between delay data and weather data sets,a piece-linear regression model that assumes weather data sets can be classified due to delay levels is proposed.Correlations analysis and error analysis are used to estimate the accuracy of this model.This paper uses the data from Qingdao Liuting airport as case study.In the case of lacking delay data,in order to classify the weather and traffic data,an improved method of using linear model and delay time attribution is proposed to decide the belongingness of data sets.As comparison,we also use the real delay data to classify the data set,verifying the effectiveness of the piece-wise linear model.