i need to find actual frequency of a bridge so that i recorded vehicle induced vibrations from an accelerometer then i processed that data in matlab by applying FFT. then i applied monte carlo simulation that randomly select the data in the form of windows and extract the frequency from that window. so the total number of windows is 30 that gave me 30 frequencies from the same vibration data. the purpose of this to filter the noise and other varibales. now the problem is i am getting different frequencies of same data no prominent frequency occured. so how can i solve this problem the scattered plot of frequency shows a particular range from 2.6 Hz to 3.1 Hz so kindly anybody tell me how can i find the actual frequency thanks in advance you can see a plot attached here thanks
I don't understand why you split up your input data.
I assume it's just ordinary time-domain accelerometer output? Why not perform your FFT on that directly?
The response will contain two sets of frequencies - one coming from the driving function that is the excitation from the vehicle and the other will be the natural frequencies. One method to identify the two would be to look for decay in magnitude from first window to last window. Natural frequency component decays while the driving function component stays on until the vehicle leaves. The range in frequency you are getting would most probably due to spectral leakage which comes due to a finite number of sample points.
- Badges
- Science topic