Discussion Forum

Yes, of course you can. This is called frequency modulation. A pure frequency-modulated waveform has a constant amplitude but its frequency can change with time (technically, this is done by changing the phase, but it has the same effect). Look up "frequency modulation" on the internet and you will find many useful references, tutorials, etc. Here is a link to one of many, many such articles: https://onlinelibrary.wiley.com/doi/pdf/10.1002/sat.4600020307 . In most cases, to represent such a waveform in Comsol Multiphysics, you will need to use a time-domain model and you should define your source waveform as a function of time.

Senior, what if I wanted to determine the excitation frequency of the next stage by the current characteristics of one cycle?

By "next stage" do you mean: (1) a later step in time, (2) some other spatial region of your problem, or (3) something else? If you are going to set your boundary conditions to be functions of FE model-computed quantities, you will need to be careful to avoid circular definitions, numerical instabilities, and more. Consider posting your model to the forum and also explaining exactly what physical process you are trying to represent with your model.