Analysis of localization errors in a mobile object
The invention relates to a method for analyzing localization errors, wherein the method includes the steps of obtaining relative positions and absolute positions of an object moving within a mapped environment, which are generated sequentially over time, calculating differences between the relative positions and absolute positions to determine sequential localization errors, and decomposing the localization errors into a systematic error that is dependent on the orientation of the object and into a systematic error that is independent on the orientation of the object by analyzing the intrinsic dynamics of the localization errors.
1. A method for analyzing localization errors, comprising the steps of:
obtaining, from position estimation modules, relative positions and absolute positions of an object moving within a mapped environment, which are generated sequentially over time;
calculating differences between the relative positions and absolute positions to determine sequential localization errors; and
decomposing the localization errors into unrelated error components by analyzing intrinsic dynamics of the localization errors.
2. The method according to claim 1 , wherein
the unrelated error components comprise a systematic error that is dependent on the orientation of the object and a systematic error that is independent of the orientation of the object.
3. The method according to claim 2 , wherein
in the obtaining step, the absolute positions are obtained from one of the position estimation modules which is an absolute position estimation module by using at least one sensor.
4. The method according to claim 1 , wherein
in the obtaining step, the relative positions are obtained from one of the position estimation modules which is a map-relative position estimation module, in which a location of the object in a map is determined; and
the method further comprises a step of correcting the map or the determined location based on the systematic error that is independent from the orientation of the object.
5. The method according to claim 3 , wherein
the method further comprises a step of calibrating at least one of the position estimation modules based on the systematic error that is dependent on the orientation of the object.
6. The method according to claim 1 , wherein
at least one of the relative positions and the absolute positions is estimated based on dead reckoning; and
the method further comprises a step of correcting the dead reckoning based on the systematic error that is dependent on the orientation of the object.
7. The method according to claim 1 , wherein
in the calculating step, difference vectors of the relative positions and absolute positions are calculated.
8. The method according to claim 7 , wherein
in the decomposing step, magnitudes and directions of the difference vectors are evaluated, wherein the magnitudes indicate error values, constant directions indicate the systematic error independent on the orientation of the object and non-constant directions indicate the systematic error dependent on the orientation of the object.
9. The method according to claim 1 , wherein
in the decomposing step, the localization errors are decomposed by a sequential estimator, in particular a recursive estimator, an Unscented Kalman Filter or a model learning algorithm.
10. An apparatus for analyzing localization errors configured to carry out the method according to claim 1 .
11. A driver assistance system comprising the apparatus according to claim 10 .
12. A vehicle comprising the driver assistance system according to claim 11 .