A new approach for implementing transfer entropy using process topology

Causal analysis is frequently applied to identify cause and effect relationships among process variables and thereby enables to obtain the propagation path of disturbances. This paper introduces a new approach to detect causality in complex systems based on the Transfer entropy (TE) method by incorporating the information on process topology using an explicit search algorithm. Initially, the TE is calculated for the pathways which are considered as direct based on the process topology. Thereafter, the Direct Transfer Entropy (DTE) is employed to discriminate spurious and/or indirect pathways. This concept is demonstrated on industrial board machine.