Acceleration
Posted: Thu Mar 09, 2017 3:05 am
Hi,
I have several questions about acceleration.
1. It seems like the data values have biases, specifically xbias = 30, ybias = -3, zbias = -83. Please, see the attached image of the histograms of the three accelerations _after_ the biases have been subtracted.
2. The positive direction of the x acceleration axis points backwards, i.e. when acceleration (after subtracting the bias) is positive the car is slowing down, and vise versa.
3. The last question is about the units. At first I assumed that, after subtracting the biases, the units are m/s^2. However, now I found the topic (viewtopic.php?f=2&t=1982&p=3936&hilit=acceleration#p3936) that states that acceleration should be computed using the following formula:
actual acceleration at X axis (in m/s^2) = x_accelerometer_sensor_data *160(to correct approximation done in original code) * 9.81(value of g)/ 16384
with some mysterious conversion factors of 160 and 16384.
What are these coefficients? And more importantly, back to my first question, is there a bias in the data?
Thank you
I have several questions about acceleration.
1. It seems like the data values have biases, specifically xbias = 30, ybias = -3, zbias = -83. Please, see the attached image of the histograms of the three accelerations _after_ the biases have been subtracted.
2. The positive direction of the x acceleration axis points backwards, i.e. when acceleration (after subtracting the bias) is positive the car is slowing down, and vise versa.
3. The last question is about the units. At first I assumed that, after subtracting the biases, the units are m/s^2. However, now I found the topic (viewtopic.php?f=2&t=1982&p=3936&hilit=acceleration#p3936) that states that acceleration should be computed using the following formula:
actual acceleration at X axis (in m/s^2) = x_accelerometer_sensor_data *160(to correct approximation done in original code) * 9.81(value of g)/ 16384
with some mysterious conversion factors of 160 and 16384.
What are these coefficients? And more importantly, back to my first question, is there a bias in the data?
Thank you