基于故障树和BP网络的飞控系统可靠性分析

Reliability analysis of aircraft control system based on FTA and BP neural network

文中针对不同批次飞行器控制系统在使用过程中的可靠度的差异问题,提出一种新的可靠度预测分析方法,该方法的思路为,首先使用故障树分析法分析出不同批次飞行器控制系统出厂时的可靠性,再根据不同批次飞行器控制系统在使用过程中的可靠度历史数据,运用BP神经网络的方法对数据进行分析处理,建立了飞行器控制系统"出厂时部件级和系统级可靠度"与"使用过程中不同年限可靠度"之间的联系与变化关系,即在统计出部件级可靠度的情况下,通过"故障树分析法"和"BP神经网络分析法"结合的方法,可以得出飞行器控制系统的可靠度,以及预测出飞行器控制系统使用过程中不同年限的可靠度。

In this paper, different batches of the aircraft control system reliability during use different problem, first use the fault tree analysis method, different batches of the factory aircraft control system reliability, according to different batches of aircraft control system in use the reliability of historical data, the use of BP neural network approach to data analysis and processing,the establishment of aircraft control system ＂factory component-level and system-level reliability＂ and ＂the process of using different term reliability,＂ the link between changes in relationship, that the statistical reliability of the component level in the case, through the ＂fault tree analysis＂ and ＂BP neural network analysis method,＂ a combination of methods, we can draw the reliability of aircraft control systems, as well as predict the aircraft control system the process of using different years of reliability.