I have spectroscopic signals that are disturbed by a pseudo-periodic signal. This disturbance is due to a bad positioning of the sensor and this one will badly integrate the number of photons it detects. Of course we could act on the positioning of the sensor, but it is mechanical, so this disturbance may still exist. So I would like to find a way to detect if there is noise or not. And have a way to model it to correct it. Can you help me ? I work with Python.
(https://i.stack.imgur.com/KKrhp.jpg)
I tried to detect the peaks, but it is not robust enough as there may be peaks that are not noise.
Here are some new signals. We see that the period and the amplitude increase.
https://i.stack.imgur.com/e2aHZ.jpg