I am measuring the thrust of an electric thruster, but the data obtained is noisy. Because the thrust value is too small (micro N), it is greatly affected by factors such as temperature, vibration, and vacuum pressure. How to denoise the thrust data and eliminate those interference items?
I think the general approach would be to measure all influences that are known to give systematic errors (you mentioned them), determine the proportionality and subtract this from the measurement. Further suppressing noise will need averaging or that combined with a lock-in approach. That means switching the thrust on and off with a (carrier-)frequency and determine the force change with this frequency in the signal. Can be done with real lock-in equipment or fourier transform of the signal.
It is possible to greatly reduce the noise level using moving averages within the equipment you use for measurements.
- Badges
- Science topic