【作者】 张君一；

【导师】 谢里阳；

【作者基本信息】 东北大学 ， 机械设计及理论， 2008， 博士

【摘要】 近年来,随机Petri网已经成为研究离散事件动态系统的一种有力的工具,特别是在制造系统可靠性评价领域取得了显著成功,关于随机Petri网在离散制造系统领域的研究越来越多。本文针对制造系统可靠性问题,按其时间分布类型分别采用广义随机Petri网(GSPN)以及扩展随机Petri网(ESPN)进行模块化建模,在此基础上进行性能分析。本文的研究结果及内容包括以下方面：1)确立制造系统的可靠性指标根据运行过程中的状态以及制造系统具体情况,引入如下相关可靠性指标：基本可靠度指标、可用度指标、维修性指标、稳态生产率；再引入功能可靠度和多任务可靠度指标,来衡量制造系统降级运行时以及多任务运行时的系统整体可靠性水平。2)采用广义随机Petri网或扩展随机Petri网针对制造系统可靠性建立模型(1)制造系统中的各种资源及其关系可以被描述成Petri网中的库所、变迁及其相应的活动,并根据变迁时延分布的不同,而分为指数与非指数分布系统,相应地采用不同的随机Petri网建模。(2)以混合建模方法建立制造系统可靠性模型,即制造系统结构自上向下建模方法,将系统结构可分为五层：工厂层、加工车间层、加工单元层、设备模块层以及设备资源层；在此基础上,采用制造资源基于模块化的自底向上的建模方法,将制造系统设备资源主要分为以下三大模块：加工设备模块、运输设备模块和仓储设备模块。3)制造系统可靠性分析针对指数分布系统,将缓冲区对加工设备生产性能的影响进行总结分类,同时对加工设备的运行状态也进行归类,通过马尔科夫状态转移方程得到其量化值,最终得到制造系统的各项可靠性指标；而针对非指数系统,根据行为表达式可以求得Petri网的传递函数,从而可以利用矩母函数思想,实现对扩展随机Petri网的性能分析。更多还原

【Abstract】 Recently, Stochastic Petri Net has become a powerful tool for the research of discrete event dynamic system and especially succeeded in the system capability evaluation field. More and more attention is paid to the application of Stochastic Petri Net in the manufacture system. With a view to analyses the problems of reliability about manufacturing systems, the whole system can be classified by times distribution, and then generalized stochastic Petri nets (GSPN) and extended stochastic Petri nets (ESPN) are adopted to modularized model respectively. At last, performance analysis will be gone on.The principal research results and contents can be outlined as following:Ⅰ) Establishing reliability measure of manufacturing systemsAccording to the states of manufacturing system and its operating, it adopts some correlation reliability indexes as follows:basic reliability index, availability index, maintainability index and stability production rate; function and multitask reliability are presented, they can measure reliability of the whole level when the system is operating in degradable or multitask.Ⅱ) Generalized stochastic Petri nets or extended stochastic Petri nets is presented to model for reliability of manufacturing systems(ⅰ) Various resources and their correlations can be described to places, transitions and corresponding behaviors of Petri nets. According to the various transition time delay distribution, system can be classified to exponential and non-exponential distribution systems, and then adopt different stochastic Petri nets to model respectively.(ⅱ) Mixed modeling method is presented for reliability model of manufacturing system. For system structure, it is built the modeling from its top to its bottom, it can be separated into five layers:plant layer, job shop layer, manufacturing cell layer, equipment module layer and machine resource layer; and for manufacturing resource, it is built the modularized modeling from its bottom to its top. it can be separated into three modules: equipment module、transport equipment module and storage module and so on. Here, reliability analysis equipment module.Ⅲ) Reliability analysis of manufacturing systemsTo systems of exponential distribution, the buffers influences to operation performance of manufacturing equipments can be classified. Similarly, it can get the classes of operation state of manufacturing equipments. And then obtain the quantization values by solving state transition equations; and to systems of non-exponential distribution, transfer functions can be obtained after getting behavior expressions. Furthermore, by using moment generating functions, realize performance analysis of model for extended generalized stochastic Petri nets.更多还原