Action planning system and method for autonomous vehicles
An action planning system ( 100 ) and method for autonomous vehicles are provided. The system ( 100 ) comprises one or more processors ( 108 ) and one or more non-transitory computer-readable storage medium ( 110 ) having stored thereon a computer program used by the one or more processors ( 108 ), wherein the computer program causes the one or more processors ( 108 ) to estimate future environment of an autonomous vehicle ( 114 ), generate a possible trajectory for the autonomous vehicle ( 114 ), predict motion and reactions of each dynamic obstacle in the future environment of the autonomous vehicle ( 114 ) based on current local traffic context, and generate a prediction iteratively over timesteps.
1. A method comprising:
operating an autonomous vehicle with one or more processors to begin a lane change action;
estimating, by the one or more processors, a future environment of the autonomous vehicle based on the lane change action;
generating, by the one or more processors, a possible trajectory for the autonomous vehicle based on the lane change action and the estimated future environment;
predicting, by the one or more processors, motion and reactions of each dynamic obstacle in the future environment of the autonomous vehicle based on current local traffic context and the possible trajectory;
iterating the predicting of the motion and reactions of each dynamic obstacle, with the one or more processors at iterative prediction timesteps;
repeating the estimating of the future environment and the generating of the possible trajectory, with the one or more processors, at action decision timesteps, which are a subset of the iterative prediction timesteps such that the repeating of the estimating of the future environment and the generating of the possible trajectory occurs less frequently than the iterating of the predicting of the motion and reactions of each dynamic obstacle; and
operating the autonomous vehicle with the one or more processors to continue the lane change action or abort the lane change action based on the predictions of the motion and reactions of each dynamic obstacle, the estimates of the future environment, and the generated possible trajectories.
2. An autonomous vehicle comprising:
one or more processors; and
one or more non-transitory computer-readable storage media having stored thereon a computer program operated by the one or more processors to:
operate the autonomous vehicle to begin a lane change action;
estimate a future environment of the autonomous vehicle based on the lane change action;
generate a possible trajectory for the autonomous vehicle based on the lane change action and the estimated future environment;
predict motion and reactions of each dynamic obstacle in the future environment of the autonomous vehicle based on current local traffic context and the possible trajectory;
iterate the prediction of the motion and reactions of each dynamic vehicle at iterative prediction timesteps;
repeat the estimating of the future environment and the generating of the possible trajectory at action decision timesteps, which are a subset of the iterative prediction timesteps such that the repeating of the estimating of the future environment and the generating of the possible trajectory occurs less frequently than the iterating of the predicting of the motion and reactions of each dynamic obstacle; and
operate the autonomous vehicle to continue the lane change action or abort the lane change action based on the predictions of the motion and reactions of each dynamic obstacle, the estimates of the future environment, and the generated possible trajectories.