Minimax Differential Dynamic Programming: An Application to Robust Biped Walking
作者： Jun Morimoto , Christopher G. Atkeson
DOI:
关键词: Robust control 、 Computer science 、 Control theory 、 Differential dynamic programming 、 State space 、 Servo 、 Servo control 、 Minimax 、 Optimal control 、 Torque
摘要: We developed a robust control policy design method in high-dimensional state space by using differential dynamic programming with minimax criterion. As an example, we applied our to simulated five link biped robot. The results show lower joint torques from the optimal compared hand-tuned PD servo controller. Results also that robot can successfully walk unknown disturbances cause controllers generated standard and fail. Learning compensate for modeling error previously conjunction is demonstrated.
nips.cc参考文章(9)
H. H. Rosenbrock, D. H. Jacobson, D. Q. Mayne, Differential Dynamic Programming The Mathematical Gazette. ,vol. 56, pp. 78- ,(1972) , 10.2307/3613752
Damien Ernst, Arthur Louette, Introduction to Reinforcement Learning MIT Press. ,(1998)
Oliver Mihatsch, Ralph Neuneier, Risk Sensitive Reinforcement Learning neural information processing systems. ,vol. 49, pp. 1031- 1037 ,(1998) , 10.1023/A:1017940631555
Peter Dyer, David M. Eggleston, Stephen R. McReynolds, The computation and theory of optimal control ,(1970)
Stefano P. Coraluppi, Steven I. Marcus, Brief Risk-sensitive and minimax control of discrete-time, finite-state Markov decision processes Automatica. ,vol. 35, pp. 301- 309 ,(1999) , 10.1016/S0005-1098(98)00153-8
Yildirim Hurmuzlu, Dan B. Marghitu, Rigid Body Collisions of Planar Kinematic Chains with Multiple Contact Points The International Journal of Robotics Research. ,vol. 13, pp. 82- 92 ,(1994) , 10.1177/027836499401300106
Jun Morimoto, Kenji Doya, Robust Reinforcement Learning neural information processing systems. ,vol. 17, pp. 1061- 1067 ,(2000) , 10.1162/0899766053011528
A.G. Barto, R.S. Sutton, Reinforcement Learning: An Introduction ,(1988)
Stefan Schaal, Christopher G. Atkeson, Constructive incremental learning from only local information Neural Computation. ,vol. 10, pp. 2047- 2084 ,(1998) , 10.1162/089976698300016963