| 摘要 |
<p>According to the method for calibrating a magnetic sensor, measurements (M I ) are acquired from a magnetic sensor (11) during a non-pre-ordered movement, and a plurality of sets of solutions (SSET K ) are determined for respective expected values (H K ) of intensity of the Earth's magnetic field. The solutions (S K ) are defined by a plurality of parameters, amongst which a gain value (GM X , GM Y , GM Z ) for each detection axis (X', Y', Z') of the magnetic sensor (11). For each solution (S K ) a figure of merit (E(S K )) is then determined, correlated to a calibration error, and a partial solution (S KB ) is selected in each set of solutions (SSET K ), on the basis of the figure of merit (E(S K )). Once a gain confidence interval (I G ) has been defined, a calibration solution (S CAL ) is selected on the basis of the figure of merit (E(S K )), from among the partial solutions (S KB ) having respective gain values (GM X , GM Y , GM Z ) all falling within the gain confidence interval (I G ).</p> |
