Ultrasonic local positioning system using active beacons, pseudorandom binary sequences correlation, CDMA & FDMA signal separation, estimation using the extended Kalman filter for synchronization, positioning refinement and data fusion with odometric measurement.
Designed for Eurobot competition at FEL CVUT http://dce.felk.cvut.cz/
Legend: bottom left plot shows correlator output, three distances from three beacons around the table; bottom right bars show odometric measurement of left/right wheel velocities, top right display shows estimated position (x, y and angle) and velocity.
In first 10s, the system has not been time synchronized, I started motion too early. After then, the estimator got locked and operated well. The erroneous measurement detector did nothing, because initial state covariance of the EKF has been too large to judge measurement as unlikely.
Link to this comment:
All Comments (0)