Using model knowledge for learning inverse dynamics

In recent years, learning models from data has become an increasingly interesting tool for robotics, as it allows straightforward and accurate model approximation. However, in most robot learning approaches, the model is learned from scratch disregarding all prior knowledge about the system. For many complex robot systems, available prior knowledge from advanced physics-based modeling techniques can entail valuable information for model learning that may result in faster learning speed, higher accuracy and better generalization. In this paper, we investigate how parametric physical models (e.g., obtained from rigid body dynamics) can be used to improve the learning performance, and, especially, how semiparametric regression methods can be applied in this context. We present two possible semiparametric regression approaches, where the knowledge of the physical model can either become part of the mean function or of the kernel in a nonparametric Gaussian process regression. We compare the learning performance of these methods first on sampled data and, subsequently, apply the obtained inverse dynamics models in tracking control on a real Barrett WAM. The results show that the semiparametric models learned with rigid body dynamics as prior outperform the standard rigid body dynamics models on real data while generalizing better for unknown parts of the state space.

[1]  Etienne Burdet,et al.  Evaluation of parametric and nonparametric nonlinear adaptive controllers , 1998, Robotica.

[2]  Jun Nakanishi,et al.  Operational Space Control: A Theoretical and Empirical Comparison , 2008, Int. J. Robotics Res..

[3]  Mark W. Spong,et al.  Robot dynamics and control , 1989 .

[4]  Jorge Nocedal,et al.  Algorithm 778: L-BFGS-B: Fortran subroutines for large-scale bound-constrained optimization , 1997, TOMS.

[5]  Dieter Fox,et al.  GP-BayesFilters: Bayesian filtering using Gaussian process prediction and observation models , 2008, 2008 IEEE/RSJ International Conference on Intelligent Robots and Systems.

[6]  Anthony Widjaja,et al.  Learning with Kernels: Support Vector Machines, Regularization, Optimization, and Beyond , 2003, IEEE Transactions on Neural Networks.

[7]  Weiping Li Adaptive control of robot motion , 1990 .

[8]  Sethu Vijayakumar,et al.  Computational Theory of Incremental and Active Learning for Optimal Generalization , 1998 .

[9]  Jun Nakanishi,et al.  A Bayesian Approach to Nonlinear Parameter Identification for Rigid Body Dynamics , 2006, Robotics: Science and Systems.

[10]  Carl E. Rasmussen,et al.  Gaussian processes for machine learning , 2005, Adaptive computation and machine learning.

[11]  Duy Nguyen-Tuong,et al.  Computed torque control with nonparametric regression models , 2008, 2008 American Control Conference.

[12]  Bernhard Schölkopf,et al.  Prior Knowledge in Support Vector Kernels , 1997, NIPS.