机载液压油源系统非马尔科夫随机Petri网模型与时间参数识别  

Non-Markovian Stochastic Petri Nets Model and Identification for Onboard Hydraulic Oil System

作　　者：刘久富[1] 陈哲[1] 王正谦[1] 王志胜[1] 杨忠[1] 

机构地区：[1]南京航空航天大学自动化学院,江苏南京210016

出　　处：《四川大学学报（工程科学版）》2013年第6期170-175,共6页Journal of Sichuan University (Engineering Science Edition)

基　　金：国家自然科学基金资助项目(60674100)

摘　　要：机载液压油源系统的正常行执行时间服从正态分布和异常行为执行时间服从指数分布。Petri网模型处理确定行为,不能处理随机行为。随机Petri网(stochastic Petri nets,SPN)模型与确定-随机Petri网(DSPN)模型处理属于指数分布的行为,不能处理正态分布的行为。提出用正态-随机Petri网(NSPN)对机载液压油源系统建模,给出正态-随机Petri网模型的正态参数与指数分布参数的识别算法。仿真计算结果表明,计算的变迁时间参数值与系统实际运行时间参数值相比误差较小,验证了该方法的有效性。The onboard hydraulic oil source system has some fault-free behaviors with normal distributed periods and some faulty ones with exponentially distributed periods. The model of the onboard hydraulic oil source system was built by normal-stochastic PNs which combined firing periods with normal and exponential distribution. A method was proposed for the the firing periods parameters identifi- cation, which was the mean value m and the periods standard deviation σ for normal distribution or the exponential distribution mean value μ. The simulation results showed that the computing operating periods parameters were consistent with the real operating periods parameters.

关 键 词：随机PETRI网 非马尔科夫 建模 参数识别 

分 类 号：TP23[自动化与计算机技术—检测技术与自动化装置]