filtering integration error

[Rap_Master_T]

Yo all,

We're working on a project with a mobile robot. We want to measure its position so we're using YEI 3-space sensor (IMU). This IMU can give us only the acceleration of the robot so we want to double integrate it and get the robot's position. In the link below we found a relevant solution for this case. However, In our understanding- there is an error caused by the double-integration which we want to minimize. Do you have any suggestion? And more specifically, Do you have a VI which can be used as a filter?

Peace,

Rap Master

https://decibel.ni.com/content/docs/DOC-5143

[MCOTO]

Hi,

I found an article in which they are using a FFT-DDI method to achieve what you are expecting (unfortunately the code is not shown). But you can use the integrated functions of LabVIEW in order to design a digital filter, depending of the version you are working.

Greetings,

[User002]

Hi,

Perhaps you should consider using a GPS or other forms of absolute-position reference systems?

Doubly-integrating will give an accumulation on the position error and sooner or later need to be "corrected" to a "true" position. 

Maybe consider dead-reckoning on the robot wheels, etc..?

https://forums.ni.com/t5/LabVIEW-Robotics-Documents/Localization-of-a-Starter-Kit-2-0-with-Dead-Reck...

https://www.ni.com/docs/en-US/bundle/labview-robotics-module/page/lvrobovi/nmea_protocol_pal.html

Br,

/Roger