Continue to Site

Eng-Tips is the largest engineering community on the Internet

Intelligent Work Forums for Engineering Professionals

  • Congratulations cowski on being selected by the Eng-Tips community for having the most helpful posts in the forums last week. Way to Go!

Calulacting localisation of an ROV with IMU 1

Status
Not open for further replies.

Dali95

Student
Nov 4, 2020
2
Hello, I am currently working on a ROV Project where I need to install a Localization system. I want to ask, is it possible to make a a sufficiently good system only with an IMU or with a speed sensor? Or do I need to add another sensor like an acoustic transducer on the surface of the water to the system. And I'm going to control it with Arduino, what's the best way to combine the results of the sensors?
 
Replies continue below

Recommended for you

"sufficiently good" is determined by your sensor accuracy, your ability to collect sensor data sufficiently fast (assuming the sensor can handle providing unique data at the desired rate), and your algorithm's (state machine?) ability to make proper use of said data. So, even with impulse inputs, a sufficiently designed system will provide valid output data... but "sufficient" is a purposely vague term.

Dan - Owner
Footwell%20Animation%20Tiny.gif
 
Using dead reckoning underwater (ie a speed sensor, clock and compass) in a slow moving vehicle in the presence of tidal currents is not going to work.


Cheers

Greg Locock


New here? Try reading these, they might help FAQ731-376
 
Thank you all for your replies.
@GregLocock yeah that what a thought too.
@MacGyverS2000 I wanted to use the adafruit bno055 do you think it's good enough the ROV can reach max. 20 meters.
@itsmoked yes that what I am looking for, I am currently looking for an acoustic transducer to add it to the system, but until now I didn't find any which can get a signal from more than approx. 30 meters and which is not expensive ( 50 euro (60 dollars)).
 
The BNO055 has a decent refresh rate, but you don't say how long you need to keep your accuracy (nor do you state what level of accuracy you need). Without knowing the error rate of Bosch's algorithm/hardware combo, along with missing info on your end, it's pretty impossible to make a useful prediction. If the ROV is going to be underwater for 5 minutes, it will likely be more than accurate for a degree or two of angle and a few inches of absolute depth... if you expect it to be there for a few hours and need sub-degree accuracy, no way, no how.

Dan - Owner
Footwell%20Animation%20Tiny.gif
 
Nice find! Cranking one's own Kalman filter to get blended attitude/position has been a big bugaboo; having a turn-key solution is pretty awesome.

Depending on the quality of the Kalman filter in the processor, its best-case orientation will be dictated by the magnetometer, since that's THE absolute orientation sensor, so ±2.5 degrees, although it might be substantially better for a well-surveyed environment, and a well calibrated sensor.

There are also maneuvers that one can do to lessen the impact of sensor drift, such as s-turns.

TTFN (ta ta for now)
I can do absolutely anything. I'm an expert! faq731-376 forum1529 Entire Forum list
 
Status
Not open for further replies.

Part and Inventory Search

Sponsor