Modeling Plant Stability in a Networked Control System

Assurance of system stability is of paramount importance in every control system. Without the maintenance of stability, plants could easily break down and explode, resulting not only in wasted time and capital, but the potential endangerment of human lives as well. This research project focuses on modeling plant stability through the emulation of a control system and the simulation of control behavior. Preparations for the investigation involve the creation of Python-scripted software that establishes the client-server relationships, automates communication between them, as well as synchronizes the timing of task delegation. The behavior of the control system is modeled by an ordinary differential equation, where a spring constant A directly affects plant stability. After the construction of the network is complete, the software is deployed on a given isolated network and tested repeatedly for stability. Results show that stability is guaranteed when the model contains a value of A between 0 and 1 for a simple two-node network. In addition to this finding, trial runs of the software were also conducted on different topologies to show an inverse relationship between information latency and stability. This paper investigates the limiting values of constant A for plant stability under different network conditions.