Aviation Risk Analysis: U-bowtie Model Based on Chance Theory

Xiaolin Wang,Lei Zhang,Jihui Xu,Mingfa Zheng

doi:10.1109/access.2019.2926210

Xiaolin Wang, Lei Zhang + Show 2 more

Open Access

https://doi.org/10.1109/access.2019.2926210

Copy DOI

Abstract

Risk analysis is a crucial component in any aviation application. The conventional risk analysis schemes for systems encompassing large amounts of data are based on the probability theory. However, in some practical systems, the components of complex systems may have few or even zero samples, in which case the risk cannot be evaluated simply via statistics. This paper proposes a novel aviation risk analysis scheme based on the chance theory, which is a generation of both probability theory and uncertainty theory. In this paper, a U-bowtie model is first established according to the graphic structure and step-wise analysis process. A risk belief metric is utilized, and the risk variable is modified by a logic algorithm based on the operational law of chance theory. Then, several risk belief formulas are built for different respective system configurations; the risk of output events in the U-bowtie model can be calculated via the corresponding formulas. Additionally, the most contributed basic event and the most effective measurement can be obtained through sensitivity analysis. The effectiveness of the proposed scheme is validated by a case study. The results presented here may represent an innovative approach to managing risk in larger classes of complex systems.

Highlights

Risk, the root cause of dangerous incidents, is a combination of the probability of occurrence of harm as well as the severity of that harm [1]–[3]
The results presented below represent an innovative approach to aviation risk analysis via U-bowtie model and closely emulate actual engineering conditions
This paper proposed an innovative aviation risk analysis approach: a U-bowtie model with a risk belief metric under which uncertainties may be disposed of effectively within realistic engineering scenarios

Summary

INTRODUCTION

The root cause of dangerous incidents, is a combination of the probability of occurrence of harm as well as the severity of that harm [1]–[3]. Wang et al.: Aviation Risk Analysis: U-bowtie Model Based on Chance Theory descriptions, including fuzzy set theory [19], [ 20] and evidence theory [21] Liu asserts that these methods may lead to counterintuitive results [22]–[23]. These two problems constitute a complex hybrid system ( called an ‘‘uncertain random system’’) characterized by randomness and indeterminacy In this case, risk quantification in bowtie analysis process cannot be conducted based only on probability theory or only on uncertainty theory. In view of the aforementioned discussion, this paper proposes a novel uncertain random bowtie (U-bowtie) risk analysis method tailored to uncertain random systems in the aviation industry based on chance theory.

UNCERTAINTY AND CHANCE THEORIES

UNCERTAIN-EVENT TREE ANALYSIS

SENSITIVITY ANALYSIS

CASE STUDY

Findings

CONCLUSION