Closed adbuerger closed 8 years ago
The problem seems to be not within the initial states, but with the definition of the measurement deviations where the wrong dimensions are used when not measuring full state. Though, the deviations are only used for setting up the objective, i. e. the actual values of the measurements are probably never used.
At the moment, a simulation results is used as an initial guess for the states within the DOE optimization problem. This fails when the measurement function differs from full states, and needs to be further investigated.