基于ANP和物元可拓模型的航空公司安全水平评价

An improved method for evaluating the safety levels of air aviation based on the ANP and the matter-element extension model

通过构建网络层次分析(ANP)和物元可拓模型相结合的评价方法,研究航空公司安全水平评价中各指标间的相互依赖、相互影响和安全水平的变化趋势。为使评价指标在实际工作中更有针对性和可操作性,根据安全管理体系(SMS)组成要素,提出了由4个准则、20项具体指标组成的航空公司安全水平评价指标体系;又因该体系中各指标间存在相互依赖、相互影响,从而引入网络层次分析法(ANP)以确定各指标权重;并与物元可拓模型结合,建立航空公司安全水平评价模型。运用该评价模型对南方某航空公司的安全水平进行评价,结果显示该模型具有可行性。

The paper is aimed at presenting an improved method for scientific and reasonable evaluation of the safety aviation level based on the study of the Analytic Network Process（ANP） method and the matter-element extension model（MEM）,in addition to the study of the interrelated and mutual influential factors likely involved as the factors to be weighed in the safety management system（SMS）. As is known,SMS has been used in many countries for quite a number of years,especially in nearly all the airline companies in China today.The evaluation system of indicating factors here we would like to put forward consists of 20 indicators of 4 categories on the basis of SMS elements method,such as the safety policy,the risk management,the safety assurance and the safety promotion. That is to say,the suggested indicator system we have worked out in this paper covers all the items for the aviation safety items illustrated in the SMS. Our system intends also to aim at taking all the indicating factors as guiding ideas for the actual aviation control work. However,there exists a problem in the above said two models,that is,the problem of the interdependence and mutual influence among all the indicators in addition to the evolutionary and developmental trends that come about with the development of the up-to-date aviation practice. To solve the said problem,we have adopted the Super Decision Software for calculating all the weights that should be taken into full account.And,the next,we have established an evaluation model by introducing into our model the ANP method and the matter-element extension model,in addition to the evaluation model in which more safety levels with the study of the developmental trends are to be included.So far,the said model of ours has been already adopted by an airline company in China south. It proves that the evaluation results we have gained come in full accord with the actual condition of the airline safety control situation. Thus,the model we have developed can be expected to offer comprehensive and reasonable evaluation in accordance with the actual need of any airline companies in order to solve any insurmountable safe management problems through the needed evaluation. The improved model for evaluating the safety level of airlines introduced in this paper can serve as a new brilliant tool and new approach to bettering the control of our air traffic.