I have a fuel cell and I want to do some system identification methods on it, my FC's input is Hydrogen and Oxygen which is obtained from water electrolysis. How can I test different input on it to measure the output voltage? Is it rational to change water electrolysis voltage? Is changing the input gas pressure the only way?
(picture attached for more guidance on question)
The test plan depends on what you want to test, many procedureres have been proposed in literature. Changing the gas pressure is not the only way, the water electrolysis voltage change has an effect on the mass flow rates, so you can change it. To identify a FC system you should know first on all the typology of FC (PEMFC, SOFC etc.), and then the test plan. For example in SOFC it is very important changing the operating temperatures, in PEMFC the umidity of membrane etc.
BR
Dr. Dario Marra
If you are new to this, then how about starting out with some simpler strategies: Current or voltage step analysis and look at the responses. It depends a bit on what equipment you have available, chirp signals or other high harmonic content signal could require some quite rapid sampling.
I would recommend using a pseudo random binary sequence, for example m-sequences of suitably long period (>127 and depending on frequency resolution required). A binary sequence is simpler to implement (on/off signal) and can obtain good frequency response results. You can check my thesis on my profile (sections 2.3.4 and 4.1-4.4)for a brief introduction to m sequence in particular.
I’m tryng to simulate and plot in matlab the RF output of a real single-side band modulator (I/Q modulation) when a chirp signal is in input.
The output of this kind of modulation when two sinusoidal tones (sin and cos) are applied to the input, is the desired sideband at f0+fc, a carrier leakage at fc and the undesired sideband leakage at f0-fc (or viceversa if the desired sideband is the lower), where f0 is the central frequency of the tx chirp and fc is the carrier frequency (local oscillator)
The RF output is calculated according to this formula
out = (I+dc1)*k*cos(wc+t+fi)+(Q+dc2)*sin(wc*t)
wher I and Q are the input components, dc1=dc2=real dc offset, causing the carrier leakage in real modulator, k=amplitude error, fi=phase imbalance error and wc=local oscillator pulsation.
my problem is that while the results are correct in the case of simple tones, in the case of chirps I can not see the sideband leakage, but there are two spectra.
- Badges
- Science topic
- Similar topics
- Computer Science
- Data Mining