Title: Performance and Optimization of Finite Closed Queueing Network Models

Speaker: James MacGregor Smith, Department of Mechanical and Industrial Engineering, University of Massachusetts, Amherst MA 01003

Time: 14:00, 11, April. (Mon.), 2016.

Venue: Room 1922, Tongji Building A . 

Abstract:

Closed finite queueing networks with finite buffers, and a fixed population circulating throughthe network are applicable to many different manufacturing and service system settings. Anovel queue decomposition approach using state dependent queues to capture the buffer of finiteM/M/c/K queues is shown to be a viable approach for modelling these systems. Each workstationis replaced with a coupled state dependent queue plus an M/M/c/K queue. An extended meanvalue analysis (MVA) algorithm is employed to demonstrate the integration of the state dependentqueues for the finite buffers in the network. Under certain restrictions concerning the network population,finite queueing networks with the state dependent queues acting as buffers should have aproduct form distribution. Several different closed series (i.e. cyclic), merge, and split topologicalsystems of finite queues are analyzed and presented. The incorporation of material handling andtransportation networks in finite buffer closed queueing networks is also studied along with theoptimization of the servers and finite buffers within the systems. 

Short Bio:

Professor Smith graduated with a B.Arch and M.Arch from the University of California atBerkeley and a Ph.D. in Mechanical and Industrial Engineering from the University of Illinois inChampaign-Urbana. Professor Smith conducts research on topological network design, stochasticnetwork design and analysis, and facility layout and location problems. In particular, he is doingresearch on Steiner minimal trees in 3d, applications of Steiner Trees to Minimum Energy Configurations(MEC’s) and protein modelling. He is also working on state dependent queueing networkanalysis and finite buffer queueing network models, quadratic assignment and set packing problems.Applications include the design and layout of manufacturing plants, health care facilities,and many other systems. One of the unique modelling tools developed in our research is concernedwith dynamic traffic flow models using queueing theory and queueing networks. He has publishedin many of the operations research journals concerned with optimization and stochastic processes.