基于Simulink的对数正态排队网络仿真及参数优化
发布时间：2018-11-11 16:50
【摘要】：随着现代科技的发展,排队论和排队网络的应用越来越广泛,从衣食住行到计算机网络,乃至航空航天都离不开排队。然而现在的排队论和排队网络,通常都假设为顾客按照泊松流到达,服务时间为指数分布。但由于现实排队的复杂多样性,这个假设具有局限性。服务时间还有可能服从对数正态分布的现象,比如维修一个可维护系统花费的时间服从对数正态分布。然而这方面的研究非常少,所以有必要对服务时间服从对数正态分布的排队模型进行研究。由于现实排队模型的复杂多样以及排队网络的不确定性,所以很难或基本不可能用精确的数学公式来表达排队模型,这时仿真是有效的解决办法。通过计算机仿真复杂的排队论或者排队网络,得出排队模型的性能参数,并为排队模型提出优化方法。本文利用MATLAB/Simulink对服务时间服从对数正态分布的排队系统和排队网络做仿真,得出不同排队系统的特点和性能参数表达式。同时通过对排队网络的仿真和分析,提出了排队网络重要的系统特征参数——等待率,根据等待率得出排队网络的瓶颈和冗余。最后通过瓶颈和冗余为优化排队网络提出解决方案。
[Abstract]:With the development of modern science and technology, queuing theory and queuing network are more and more widely used, from food, clothing, housing and transportation to computer network, even aviation and aerospace can not do without queuing. However, the current queuing theory and queuing network usually assume that customers arrive according to Poisson flow and service time is exponentially distributed. However, due to the complexity and diversity of the real queue, this assumption has its limitations. It is also possible to use the logarithmic normal distribution of service time, such as the time taken to maintain a maintainable system from the logarithmic normal distribution. However, there is very little research on this aspect, so it is necessary to study the queuing model of service time service from logarithmic normal distribution. Because of the complexity and diversity of the real queuing model and the uncertainty of the queuing network, it is difficult or impossible to express the queuing model with accurate mathematical formulas. At this time, simulation is an effective solution. Through computer simulation of complex queuing theory or queuing network, the performance parameters of the queuing model are obtained, and the optimization method for the queuing model is proposed. In this paper, MATLAB/Simulink is used to simulate the queuing system and queuing network with logarithmic normal distribution, and the characteristics and performance parameter expressions of different queuing systems are obtained. At the same time, through the simulation and analysis of queueing network, the important characteristic parameter of queueing network, waiting rate, is put forward, and the bottleneck and redundancy of queueing network are obtained according to the waiting rate. Finally, a solution for optimizing queueing network is proposed by bottleneck and redundancy.
【学位授予单位】：大连海事大学
【学位级别】：硕士
【学位授予年份】：2015
【分类号】：O226
本文编号：2325510
[Abstract]:With the development of modern science and technology, queuing theory and queuing network are more and more widely used, from food, clothing, housing and transportation to computer network, even aviation and aerospace can not do without queuing. However, the current queuing theory and queuing network usually assume that customers arrive according to Poisson flow and service time is exponentially distributed. However, due to the complexity and diversity of the real queue, this assumption has its limitations. It is also possible to use the logarithmic normal distribution of service time, such as the time taken to maintain a maintainable system from the logarithmic normal distribution. However, there is very little research on this aspect, so it is necessary to study the queuing model of service time service from logarithmic normal distribution. Because of the complexity and diversity of the real queuing model and the uncertainty of the queuing network, it is difficult or impossible to express the queuing model with accurate mathematical formulas. At this time, simulation is an effective solution. Through computer simulation of complex queuing theory or queuing network, the performance parameters of the queuing model are obtained, and the optimization method for the queuing model is proposed. In this paper, MATLAB/Simulink is used to simulate the queuing system and queuing network with logarithmic normal distribution, and the characteristics and performance parameter expressions of different queuing systems are obtained. At the same time, through the simulation and analysis of queueing network, the important characteristic parameter of queueing network, waiting rate, is put forward, and the bottleneck and redundancy of queueing network are obtained according to the waiting rate. Finally, a solution for optimizing queueing network is proposed by bottleneck and redundancy.
【学位授予单位】：大连海事大学
【学位级别】：硕士
【学位授予年份】：2015
【分类号】：O226
【参考文献】
相关期刊论文 前1条
1 王建江;邱涤珊;彭黎;;基于排队网络的空间信息数据处理系统效能评估[J];计算机应用;2012年03期
相关博士学位论文 前1条
1 周宗好;通信网络中的排队模型研究[D];江苏大学;2011年
相关硕士学位论文 前2条
1 吴江霞;正态分布进入统计学的历史演化[D];河北师范大学;2008年
2 曲超;集装箱码头物流资源配置仿真研究[D];山东大学;2012年
，本文编号：2325510
本文链接：https://www.wllwen.com/kejilunwen/yysx/2325510.html
