Why my genetic algorithm gives variables with unlogical dimensions?

Hi friends. I use genetic algorithm to minimize the difference between two signals(one a real signal and the other is a signal obtained from running a finite element model multiple times). I want GA algorithm to find some parameters so that these two signals match.

The cost function is the mean square error of Instantaneous frequency of signals(using Hilbert Transform). the instantaneous frequcy dimensions are not correct and also,they differ for the two signals.

any idea why this happens?

Thanks in advance