基于区间数特征根法的民航事故征候诱因分析

Research on Inducement Index and Weight Determination of Civil Aviation Accident Symptom

评价指标对模糊综合评价的结果准确性具有直接影响,民航事故征候诱因指标构建的科学性对民航安全评价具有重要现实意义,但传统层次分析法存在对专家经验要求高、主观因素对评价结果影响大等局限,区间数特征根法(IEM)可弥补其不足。首先运用SPSS软件对民航事故征候诱因指标进行相关性和因子分析,得到了指标之间关联性特征,然后运用MATLAB软件构造IEM程序,得出影响民航事故征候诱因指标权重。结果表明:皮尔逊相关系数法、因子分析法、IEM的联合使用可减少评价体系的冗余度,为指标和权重确定的合理性奠定理论基础。

The evaluation index have a direct impact on the accuracy of the results of the fuzzy comprehensive evaluation.The scientificity of the construction of the civil aviation accident inducement index have important practical significance for the civil aviation safety evaluation.However,the traditional AHP has high requirements for expert experience and subjective factors affect the evaluation results.Due to the limitations of large and other areas,the interval eigenvalue method(IEM)can make up for its shortcomings.Firstly,SPSS software is used to analyze the correlation and factor of the causative indicators of civil aviation incidents,and the correlation characteristics between the indicators are obtained.Then,the IEM program is constructed by MATLAB software to obtain the weights of the causative indicators affecting civil aviation incidents.The results show that the combined use of Pearson’s correlation coefficient method,factor analysis method and IEM can reduce the redundancy of the evaluation system and lay a theoretical foundation for the rationality of the determination of indicators and weights.