A Queueing Network Approach for Modelling Control Nets in Building Automation

Abstract:
We present a queueing network approach to build and analyze models of control nets in building and factory automation. The queueing network models are generated automatically from the design database of the control network, only a few load parameters have to be added such that model generation is extremely efficient. The resulting models are open multi-class queueing networks including finite capacity FCFS stations and a specific station type to model network access in CSMA-based control networks. The model class and approximate solution techniques are introduced. Analysis of the models is done by a fixed point approach where in each solution step an approximate product form algorithm is applied. The algorithm is very efficient and can be used to analyze realistic scenarios with several thousand stations in a few seconds. Furthermore, it is briefly outlined how the queueing network analysis is integrated in a tool environment to design LON-based control networks in buildings.