Direct multiple shooting for nonlinear optimum experimental design
Optimum experimental design (OED) for parameter identification has become a key technique in the model validation process for dynamical systems. This paper deals with optimum experimental design for systems modelled by differential-algebraic equations. We show how to formulate OED as a nonstandard n...
Saved in:
| Main Authors: | , , |
|---|---|
| Format: | Chapter/Article Conference Paper |
| Language: | English |
| Published: |
2015
|
| In: |
Multiple Shooting and Time Domain Decomposition Methods
Year: 2015, Pages: 115-141 |
| DOI: | 10.1007/978-3-319-23321-5_4 |
| Subjects: | |
| Online Access: | Verlag, Volltext: http://dx.doi.org/10.1007/978-3-319-23321-5_4 Verlag, Volltext: https://link.springer.com/chapter/10.1007/978-3-319-23321-5_4 
|
| Author Notes: | Dennis Janka, Stefan Körkel, Hans Georg Bock |
| Summary: | Optimum experimental design (OED) for parameter identification has become a key technique in the model validation process for dynamical systems. This paper deals with optimum experimental design for systems modelled by differential-algebraic equations. We show how to formulate OED as a nonstandard nonlinear optimal control problem. The direct multiple shooting method is a state of the art method for the solution of standard optimal control problems that leads to structured nonlinear programs. We present two possibilities how to adapt direct multiple shooting to OED by introducing additional variables and constraints. We highlight special structures in the constraint and objective derivatives whose evaluation is usually the bottleneck when solving dynamic optimization problems by multiple shooting. We have implemented a structure exploiting algorithm that takes all these structures into account. Two benchmark examples show the efficiency of the new algorithm. |
|---|---|
| Item Description: | Gesehen am 30.01.2018 |
| Physical Description: | Online Resource |
| ISBN: | 9783319233215 |
| DOI: | 10.1007/978-3-319-23321-5_4 |