论文信息 - Fast Model Predictive Control for Reactive Robotic Swimming

Fast Model Predictive Control for Reactive Robotic Swimming

Model Predictive Control amounts to re-solving a trajectory optimization problem at every timestep. Using a fast multi-threaded simulator and an efficient trajectory optimizer, we generate complex, reactive and robust swimming behavior in simulated robots of 12-27 state dimensions, with time-steps in the range of 10-40ms.

Yuval Tassa

[1] M. L. Chambers. The Mathematical Theory of Optimal Processes , 1965 .

[2] David Q. Mayne,et al. Differential dynamic programming , 1972, The Mathematical Gazette.

[3] L. Liao,et al. Convergence in unconstrained discrete-time differential dynamic programming , 1991 .

[4] L. Liao,et al. Advantages of Differential Dynamic Programming Over Newton''s Method for Discrete-time Optimal Control Problems , 1992 .

[5] Karl Sims,et al. Evolving virtual creatures , 1994, SIGGRAPH.

[6] M. Diehl,et al. Real-time optimization and nonlinear model predictive control of processes governed by differential-algebraic equations , 2000 .

[7] Jun Morimoto,et al. Minimax differential dynamic programming: application to a biped walking robot , 2003, SICE 2003 Annual Conference (IEEE Cat. No.03TH8734).

[8] E. Todorov,et al. A generalized iterative LQG method for locally-optimal feedback control of constrained nonlinear stochastic systems , 2005, Proceedings of the 2005, American Control Conference, 2005..

[9] Pieter Abbeel,et al. An Application of Reinforcement Learning to Aerobatic Helicopter Flight , 2006, NIPS.

[10] Hans Joachim Ferreau,et al. Efficient Numerical Methods for Nonlinear MPC and Moving Horizon Estimation , 2009 .