Bounded-Error LQR-Trees
Publication
, Conference
Ames, B; Konidaris, G
Published in: IEEE International Conference on Intelligent Robots and Systems
November 1, 2019
We present a feedback motion planning algorithm, Bounded-Error LQR-Trees, that leverages reinforcement learning theory to find a policy with a bounded amount of error. The algorithm composes locally valid linear-quadratic regulators (LQR) into a nonlinear controller, similar to how LQR-Trees constructs its policy, but minimizes the cost of the constructed policy by minimizing the Bellman Residual, which is estimated in the overlapping regions of LQR controllers. We prove a sample-based upper bound on the true Bellman Residual, and demonstrate a five-fold reduction in cost over previous methods on a simple underactuated nonlinear system.
Duke Scholars
Published In
IEEE International Conference on Intelligent Robots and Systems
DOI
EISSN
2153-0866
ISSN
2153-0858
ISBN
9781728140049
Publication Date
November 1, 2019
Start / End Page
144 / 150
Citation
APA
Chicago
ICMJE
MLA
NLM
Ames, B., & Konidaris, G. (2019). Bounded-Error LQR-Trees. In IEEE International Conference on Intelligent Robots and Systems (pp. 144–150). https://doi.org/10.1109/IROS40897.2019.8967750
Ames, B., and G. Konidaris. “Bounded-Error LQR-Trees.” In IEEE International Conference on Intelligent Robots and Systems, 144–50, 2019. https://doi.org/10.1109/IROS40897.2019.8967750.
Ames B, Konidaris G. Bounded-Error LQR-Trees. In: IEEE International Conference on Intelligent Robots and Systems. 2019. p. 144–50.
Ames, B., and G. Konidaris. “Bounded-Error LQR-Trees.” IEEE International Conference on Intelligent Robots and Systems, 2019, pp. 144–50. Scopus, doi:10.1109/IROS40897.2019.8967750.
Ames B, Konidaris G. Bounded-Error LQR-Trees. IEEE International Conference on Intelligent Robots and Systems. 2019. p. 144–150.
Published In
IEEE International Conference on Intelligent Robots and Systems
DOI
EISSN
2153-0866
ISSN
2153-0858
ISBN
9781728140049
Publication Date
November 1, 2019
Start / End Page
144 / 150