BP神经网络在航空机务人员本质安全程度评价中的应用

航空机务是飞机安全服役的重要保障,为提高航空机务系统的本质安全性,需要对航空机务人员的本质安全程度进行准确评价。在对本质安全人进行定义并明确航空机务人员本质安全程度评价项目后,构建了一种基于MATLAB的用于本质安全程度评价的BP神经网络模型。实例分析中,以某航空公司10位机务人员的专家打分数据作为样本输入,在对输入数据进行标准化处理,并明确期望输出后,通过编程计算,确定了网络隐含层神经元的最佳数目。采用优化结构进行仿真计算,结果表明BP网络的期望输出值和实际仿真输出值能较好吻合,证实了模型的可信性。基于BP网络的航空机务人员本质安全程度评价具有较好的适用性和可行性。

Application of BP neural network in inherent safety degree evaluation of aviation maintenance personnel

Aviation maintenance is the important foundation to aircraft service safety.In order to improve the inherent safety of aviation maintenance system,the exact evaluation of the degree of aviation maintenance personnel is necessary.After the definition of inherent safety human and the nail down of evaluation index for inherent safety degree of aviation maintenance personnel,a BP neural network based on MATLAB was built to evaluate inherent safety degree.In the example analysis,by using the expert mark data of a aviation company as the input data,after normal dealing with input data and defining the expect output,the best nerve cell numbers of connotative layer was confirmed by programming and computation.The optimal structure was applied to do the simulation computation.The simulation results showed that net expect results matched well with the simulation results,which validate the creditability of the BP model.Evaluation of aviation maintenance personnel inherent safety degree based on BP network has preferable applicability and feasibility.