A partial Granger causality based method for analysis of parameter interactions in bioreactors

In this article, we utilize the concept of partial Granger causality to study the penicillin production process under several operating conditions. We propose a graph-theoretic template (causal network) based method for intelligent process monitoring. We validate our results with the aid of existing knowledge and available literature. The proposed method is quite general and can be extended to analyze several physical, chemical or biological systems.