Cooperative aircraft navigation
A UAS may use camera-captured ground feature data to iteratively estimate pose change through a Kalman filter. To avoid excessive drift, the UAS may periodically perform a reset by publishing delta pose and covariance as an edge in a global pose graph, by zeroing out the delta pose and covariance. To further improve global pose estimation, two UASs may share range information and associated pose graph information, thereby providing an additional constraint for each UAS to use in its pose graph optimization.
1. A method for improving navigation of an aircraft system, comprising:
receiving, from one or more sensors, sensor data reflecting movement of the aircraft system relative to one or more ground features;
estimating, based on the sensor data, and using a relative odometry model, at least one of the following: an estimated change in x,y position since an earlier reset time, and an estimated change in yaw angle since an earlier reset time;
determining, based on a reset criteria, to reset the relative odometry model, wherein resetting the relative odometry model comprises at least one from the following list:
setting the estimated change in x,y position since the earlier reset time to zero; and
setting the estimated change in yaw angle since the earlier reset time to zero;
publishing to a global odometry model a relative odometry update comprising at least one of the following: the estimated change in x,y position since the earlier reset time, and the estimated change in yaw angle since the earlier reset time;
resetting the relative odometry model;
incorporating the relative odometry update into the global odometry model as a constraint on at least one of the following: global x,y position and global yaw angle;
estimating global pose information in the global odometry model by applying nonlinear unconstrained optimization; and
using two or more global constraints to calculate scale bias and to distribute a scale bias correction over at least two relative odometry constraints.
2. A method for improving navigation of an aircraft system, comprising:
receiving, from one or more sensors, sensor data reflecting movement of the aircraft system relative to one or more ground features;
estimating, based on the sensor data, and using a relative odometry model, at least one of the following: an estimated change in x,y position since an earlier reset time, and an estimated change in yaw angle since an earlier reset time;
determining, based on a reset criteria, to reset the relative odometry model, wherein resetting the relative odometry model comprises at least one from the following list:
setting the estimated change in x,y position since the earlier reset time to zero; and
setting the estimated change in yaw angle since the earlier reset time to zero;
publishing to a global odometry model a relative odometry update comprising at least one of the following: the estimated change in x,y position since the earlier reset time, and the estimated change in yaw angle since the earlier reset time;
resetting the relative odometry model;
incorporating the relative odometry update into the global odometry model as a constraint on at least one of the following: global x,y position and global yaw angle;
estimating global pose information in the global odometry model by applying nonlinear unconstrained optimization; and
using two or more global constraints to calculate scale bias and to distribute a scale bias correction over at least two relative odometry constraints;
wherein the global odometry model comprises:
one or more relative odometry constraints based on one or more relative odometry updates; and
one or more estimates of global x,y position and/or yaw angle based on the one or more relative odometry constraints; and
further comprising:
simultaneously:
obtaining a range to a second aircraft having a second-aircraft global odometry model;
sharing the range with the second aircraft;
resetting the relative odometry model and publishing a relative odometry update to the global odometry model;
receiving, from the second-aircraft global odometry model, information for linking the global odometry model to the second-aircraft odometry model; and
based on the information received from the second-aircraft global odometry model, adjusting the global odometry model.
3. The method of claim 2 , further comprising incorporating into the global odometry model the entirety of the second aircraft's global odometry model.
4. The method of claim 2 , comprising receiving, from the second aircraft, a segment of a third aircraft's global odometry model, and using the segment of the third aircraft's odometry model to adjust the global odometry model.
5. A method for improving navigation of an aircraft system, comprising:
maintaining a global odometry model comprising a pose graph, the pose graph comprising delta-pose edges and global-pose-estimate nodes;
simultaneously:
obtaining a range to a second aircraft having a global odometry model;
sharing the range with the second aircraft;
resetting a relative odometry model and publishing a relative odometry update to the global odometry model;
receiving, from the second aircraft's global odometry model, information for linking the global odometry model to the second aircraft's odometry model;
based on the information received from the second aircraft's global odometry model, adjusting the global odometry model
estimating global pose information in the global odometry model by applying nonlinear unconstrained optimization; and
using two or more global constraints to calculate scale bias and to distribute a scale bias correction over at least two relative odometry constraints.