A supervised approach towards network control system modelling

Abstract

Over the last decades, major advancements in wired and wireless communication networks gave rise to the Networked Control Systems (NCS) for bringing automation to traditional control systems. Different components of a control system interact with each other through a network without any human intervention. Traditional control system models is not very suitable for predicting the stability of the system. In this paper, we propose a supervised approach for modelling a NCS using the previous working data of the system. The supervised model can be either a classifier or a regression model based on whether the output to be predicted is discrete or continuous. The main advantage of our approach is the ability to use it on any NCS provided that the previous working data is available. We have used dimensionality reduction techniques for dealing with large amount of features. The experimental results showcase the efficiency of the proposed approach.