An Approach to the Calibration of Modelica Models
Miguel A. Rubio, Alfonso Urquía, S. Dormido · 2007
Abstract. An approach to the calibration of Modelica models using genetic algorithms (GA) is presented. The functions required to perform the model calibration have been programmed in the Modelica language and structured in a Modelica library, called GAPILib. This Modelica library is intended for parameter estimation in any Modelica model, supporting simple-objective optimization. Model calibration with GAPILib does not require to perform model modifications. During the algorithm run, the user can interactively change the value of the GA parameters. In addition, GAPILib supports parameter sensitivity analysis, and it is well suited for parallel computing. GAPILib is a free library (available on