The vehicle is moving at constant average speed of 20Km/hour. The vibration due to road profile is logged using tri-axial accelerometer and filtered to remove noise. The cutoff frequency is the question. The sampling rate is 20Hz.
thanks mohamed el badaoui. i am using chebyshev filter and it uses cutoff frequency in range 0 and 1. How to get it? should 8Hz divide by sampling frequency?
thanks
I am using mat lab cheby1 function. I have acceleration signal captured over time.
The device constr?aints to keep. Could you please elaborate
sir
20 Hz sampling rate is too low. I agree with Mohamed_El_badaoui. You should increase the sampling rate. I would suggest 500 to 1000 Hz. Then choose the LP filter considering the accelerations you want to measure. I suggest you should measure accelerations up to 10 g.
Fernando Thanks for your answer. I am not using special accelerometer device. I am using Android device and the app is limited for logging frequency of 20Hz as increasing it may cause problem in logging. Please suggest what to do if the limitations to keep.
Rajiv
Within those limitations the maximum frequency you can measure is 10 Hz. You need to study the software you are using. Your 0 and 1 limits are probably 0 Hz and Fsam/2, that is, 10 Hz. You might want to choose 8 Hz as your maximum, as Mohamed El Babaoui suggested, so the argument is = 8/10 = 0.8.
But do check if what I assumed here is correct, that is, that your 1 as the maximum argument corresponds to Half of the sampling frequency.
Ah, and do attach your android to the chassis of the car, do not hold it in your hand as this will filter high frequencies!!
Have fun! :-)
Thanks Fernando for your valuable input.
I want to keep android device as usually it is handled either in pocket or on car seat to give natural handling environment. I am trying it in areas where cars cannot reach with device in pocket of shirt. It is not tied or fixed anywhere. The limitations like damping due to human body and other factors is to keep. Can you suggest something on it?
Rajiv
If you want to measure vehicle vibration I would again strongly suggest you do NOT use the device in your pocket or your hands, but attach it to the vehicle (even buuble gum would do). Try the 3 situations and see for yourself what happens. Ah and I checked 'help cheby1' in Matlab and what I said before is correct, that is, the normalised range [0, 1] for Wn corresponds to [0, Fs/2].
Best of luck.
vehicle vibration measured using accelerometer sensor fixing must . so it should do attach tie are something your vechile.but you are using android it is not for industrial use on test tracks
measured vibration It 's be the relative movement of various vehicle parts.noted that in the ground transportation acting forces are of primary interest, As a result, the preferred measurand is the absolute acceleration followed by the relative displacement.
It is experience that the vibratory acceleration avobserved in the ground transportation is smaller than standard gravity acceleration gN ≈9.81 m s-2
.
- Badges
- Science topic
- Similar topics
- Engineering
- Electrical Engineering
- Signal Processing