search

FS-Driverless / Formula-Student-Driverless-Simulator

A virtual world where Autonomous Systems from different Formula Student teams can compete in time-trial challenges
https://fs-driverless.github.io/Formula-Student-Driverless-Simulator
GNU General Public License v2.0
210 stars 91 forks source link

IMU Sensor Noise #134

Open wahllca opened 4 years ago

wahllca commented 4 years ago

Hi, is there any way to reduce the noise of the IMU Sensors in the Simulation? The data below was recorded during a straight acceleration, and as you can see, the noise is quite high (~80% of scale) compared to real IMU data and therefore almost useless.

dataSensor

SijmenHuizenga commented 4 years ago

@davidoort @skimslozo @CharlieA0 what do you think about this?

skimslozo commented 4 years ago

@wahllca The current IMU noise parameters are modelled according to a real IMU (see datasheet). How do the noise parameters of your mentioned reference IMU compare to this one?

wahllca commented 4 years ago

Thank you very much for the quick response! Maybe I am a little spoiled with our self-developed IMU, but you are referring to a rather cheap IMU compared to the prices of other sensors of the autonomous system. Maybe we can agree on the Sparkfun Razor IMU 9DOF (http://wiki.ros.org/razor_imu_9dof) because it is a widely used device. I could also provide a recorded rosbag for reference.

SijmenHuizenga commented 4 years ago

I don't mind moving to a better imu model at all! It seems like a small change and I can't imagine other teams hate having access to better hardware.

@wahllca Would you be able to propose changes to this file?

SijmenHuizenga commented 4 years ago

@skimslozo Do you think this is an easy change? If so, would you be able to implement this?

skimslozo commented 4 years ago

@SijmenHuizenga Yes, will take a look.