Implement support for fusing delayed measurements by keeping a state and measurement history #266
Conversation
Post-linting merge
Merging orientation Mahalanobis fix
navsat_transform GPS offset improvements
… to comply with ROS standard.
|
Will review this soon. Thanks for submitting! |
|
I'll try to give it a spin shortly as well. Thanks! |
| void validateDelta(double &delta); | ||
|
|
||
| protected: | ||
|
|
There was a problem hiding this comment.
Please remove these blank lines.
| const FilterState &old = filterStateHistory_.back(); | ||
| filter_.setState(old.state_); | ||
| filter_.setEstimateErrorCovariance(old.estimate_error_covariance_); | ||
| filter_.setLastMeasurementTime(old.time_); |
There was a problem hiding this comment.
Without looking deeper (I will), do we need to restore the filter's lastUpdateTime as well?
|
Thanks for the PR! We can iterate on the potential issues I've found. Would you happen to have a bag file that can be used to test this? I should probably add an integration test for this. |
|
Hi, |
|
No rush! Thanks for your help. |
|
FYI, I addressed the issues I pointed out. I changed the names of the parameters; I'm open to other PRs to change the names if we feel they're not representative of what the new feature does. Thanks again for the PR! |
3 participants
This implements a smoothing scheme as described in issue 263 and also adresses the issue 262.