Multidisciplinary Analysis and Optimization

Parameter Identification

After modeling process, which considers sufficient all physical effects and its interactions, some model parameters have to be validated. Finding of the right model parameters depends whether the model behavior reflects the system response accurately. It is possible in OptiY to find the required model parameters by defining a optimization task. A evaluation signal has to define as integral difference between the measure signal and the target signal during simulation:

evaluation signal = Integral (|measure signal - target signal|)

The criterion of the optimization problem should be accumulated from this evaluation signal. Within, the surface between measuring signal and target signal has to be minimized by optimization to coincide the both curves.

In OptiY, a novel optimization technology has been developed based on 1D meta-modeling. The parameter identification can be done extremely fast. With this technology, computing intensive dynamical simulation can be also optimized in available time.

 

      

before optimization                                                             after optimization