Data-driven causality digraph modeling of large-scale complex system based on transfer entropy

Standard formal mathematical tools are usually not suitable to address the problem of large scale-systems modeling since the differential equations describing the behavior of such systems are very difficult to obtain. This is due, in particular, to the fact that the analytical representation of the underlying physical laws is often unknown or is too complex for numerical considerations. In this paper, we propose a method, based on transfer entropy analysis, to identify the causal relationships between process measured variables in order to have a digraph model. This graphical model can be used in root cause and hazard propagation analysis. A case study based on three-tanks system is presented to illustrate the application of the proposed methods.