LAUSR

Abstract In this contribution, a novel method for model order reduction is presented. Since biological systems are often highly nonlinear or having a high number of parameters, the concept of empirical Gramians is considered for controllability and observability analysis. This allows a balanced representation of the system, which offers model reduction by truncation. We motivate to transfer the idea of balancing states to of balancing parameters. For that, loadability Gramian is introduced first, characterizing controllability Gramian for parameters. Second, parameters were treated as constant states, which allows observability-based identification. It is then possible to obtain a balanced realization of the parameter-space of the system, which enables balanced truncation. This method is applied to systems of glucose-insulin homeostasis.