A type-2 fuzzy modelling framework for aircraft taxi-time prediction

Knowing aircraft taxi-time precisely a-priori is increasingly important for any airport management system. This work presents a new approach for estimating and characterising the taxi-time of an aircraft based on historical information. The approach makes use of the interval type-2 fuzzy logic system, which provides more robustness and accuracy than the conventional type-1 fuzzy system. To compensate for erroneous modelling assumptions, the error distribution of the model is further analysed and an error compensation strategy is developed. Results, when tested on a real data set for Manchester Airport (U.K.), show improved taxi-time accuracy and generalisation capability over a wide range of modelling assumptions when compared with existing fuzzy systems and linear regression-based methods.