Nonlinear Parameter Identification for Ordinary Differential Equations

In many applications with dynamical processes parameter depending systems of nonlinear ordinary differential equations (ODE) are formulated. Often the model description does not perfectly match with realistic data. SQP-methods can be used to identify the parameters of the dynamical model. In conventional approaches the ODE system is solved numerically several times during each of the iteration steps of the optimization process. In this work it is proposed to perform the numerical integration of the ODE system within the optimization process. The benefits of the presented technique will be illustrated by the application of a turbocharger design within the context of diesel engined vehicle development. (© 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim)