Error Analysis and Interval Prediction of Aviation Safety Prediction Based on Uncertainty

Abstract

Aviation safety prediction is mostly deterministic prediction, which ignores the influence of various uncertainties on the prediction results. In order to quantitatively measure the contributions of the uncertainty of forecast error to the prediction of aviation safety, a novel method, which can obtain the prediction of aviation safety interval considering the uncertainty of error, the properties of which are also discussed. Meanwhile, this method aim at determining the area contains between aviation safety forecast reliability, which better understand be predicted quantity change in the future of the uncertainty and risk. Then, Taking aviation safety data of civil aviation from 1994 to 2015 as an example, the results show that the proposed aviation safety interval prediction can provide aviation safety prediction curve and variation range of this curve, which is more conducive to modeling uncertainty of aviation safety.