模糊专家系统的连锁失效方法及在船舶消防中的应用

本文提出了一种基于模糊专家系统的连锁性失效测评方法,建立了基于该系统的连锁性失效测评指标,该方法不同于以往故障树方法的静态性,具有实时动态性,根据不同的输入参数,能够给出不同的决策结果,更能满足实际的需要。文章以船舶火灾自动报警系统为例,对船舶火灾自动报警系统以舱室为单元进行模块化建模,考虑了温度、烟雾浓度、可燃气体浓度,舱室危险度尤其是舱室之间关联信息、报警时间和舱室着火温度三个对连锁失效起重要作用的影响因子,详细设计了模糊规则,确定连锁报警舱室,判定连锁报警等级,得到船舶舱室消防系统单元连锁性失效路径。仿真结果说明了该方法的准确性与有效性,表明测评结果的准确性。最后,通过和故障树方法对比阐述了该方法的优越性。

Cascading Failure Evaluation of Ship Automatic Fire Alarm System Based on Fuzzy Expert System

The cascading failure evaluation method based on fuzzy expert system and the cascading failure index were proposed in this paper. Different from the static properties of fault tree analysis, the method proposed possess the real-time dynamic and have the ability to make the corresponding strategic decisions according to different input parameters, which can better meet the actual needs. This paper took the ship automatic fire alarm system as an example, modular modeling it with a cabin as a unit. Temperature, smoke density, combustible gas density, cabin hazard, especially the link between cabins, alarm time and temperature of fired cabin, three most important factors, were considered. The fuzzy rules were detailed designed, which could determined the cascading alarm cabins, confirm the grade of cascading alarm, and achieved the cascading failure path of ship cabin units of fire-fighting system. Accuracy and validity of the method proposed were proved by the simulation results, indicating the accuracy of evaluation. Finally, advantages were interpreted by the contradistinction with the fault tree method.