2021-01-29, 09:23 PM
The sensors magnetic sensors on most of the imu are not very good. The precision is often much better than the accuracy. To read the correct compass angle it must be tilt compensated which needs other sensors, and on a moving platform accelerometers aren't enough for this.
So gyros are the simplest sensor to get the needed correction since it can use well understood algorithms like kalman filters. They also happen to give a much better resolution of angular rate than differentiating the compass without the introduced errors of tilt compensation.
To get the same performance without gyros is possible using machine learning and maybe a video camera looking at the waves and much more processing power. I'm trying to figure out how to do this anyway but I will keep the gyros, and maybe even add a few more (redundant sensors) mounted in different orientations to reduce noise and detect magnetic distortions and also automatically calibrate the accelerometers and gyros for scale factors and non linearities and temperature coefficients and mis alignments on the fly thus reducing the errors from 1-2 degrees down to 0.5 or better...
So gyros are the simplest sensor to get the needed correction since it can use well understood algorithms like kalman filters. They also happen to give a much better resolution of angular rate than differentiating the compass without the introduced errors of tilt compensation.
To get the same performance without gyros is possible using machine learning and maybe a video camera looking at the waves and much more processing power. I'm trying to figure out how to do this anyway but I will keep the gyros, and maybe even add a few more (redundant sensors) mounted in different orientations to reduce noise and detect magnetic distortions and also automatically calibrate the accelerometers and gyros for scale factors and non linearities and temperature coefficients and mis alignments on the fly thus reducing the errors from 1-2 degrees down to 0.5 or better...