I am currently utilizing the "imu_complimentary_filter" to estimate a drone's attitude. During my use of the filter, I encountered difficulties with slow attitude movement estimations. However, I discovered that disabling the bias estimation significantly improved performance.
Given my use-case scenario, the drone frequently remains stationary for extended periods before and between flights. Hence, having the ability to dynamically alter bias estimation after the node has been launched would be highly beneficial.
Is this anything that you would be interested to have upstream, if so, how would you like me to implement it?
Fist thought was to either make a ros service or to make a separate timer that checks if the parameter has been changed, and accordingly update the filter_.do_bias_estimationmember.
Could of course also be a subscriber listening to a specific topic.
Hi,
I am currently utilizing the "imu_complimentary_filter" to estimate a drone's attitude. During my use of the filter, I encountered difficulties with slow attitude movement estimations. However, I discovered that disabling the bias estimation significantly improved performance.
Given my use-case scenario, the drone frequently remains stationary for extended periods before and between flights. Hence, having the ability to dynamically alter bias estimation after the node has been launched would be highly beneficial.
Is this anything that you would be interested to have upstream, if so, how would you like me to implement it?
Fist thought was to either make a ros service or to make a separate timer that checks if the parameter has been changed, and accordingly update the
filter_.do_bias_estimationmember.Could of course also be a subscriber listening to a specific topic.
Any thoughts on this?