Dynamics systems vs. optimal control--a unifying view.

By Stefan Schaal, Peyman Mohajerian, Auke Jan Ijspeert. 2007

284 citations recorded by the source corpus.

Subfields: Biomedical Engineering, Cell Biology, Cognitive Neuroscience

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Biomedical Engineering

In Cell Biology

In Cognitive Neuroscience

Citations in the corpus

Reinforcement learning in robotics: A survey (2013, background): “Here, wi contains the ith adjustable parameter of all DoFs, N is the number of parameters per DoF, and ψi( z) are the corresponding weighting functions (Schaal et al., 2007).”
Action understanding and active inference (2011, unknown): “Conceptually, they can be thought of as encoding dynamical movement ‘primitives’ (Ijspeert et al. 2002; Schaal et al. 2007; Namikawa and Tani 2010) or perceptual and motor ‘schema’ (Jeannerod et al.”
Learning position and orientation dynamics from demonstrations via contraction analysis (2018, background): “In Schaal et al. (2007), a comparison between dynamical system-based methods and optimal control meth-ods is provided.”
Learning motor primitives for robotics (2009, methodology): “could be shown that a single first order system suffices [4]”
Learning Partially Contracting Dynamical Systems from Demonstrations (2017, unknown): “In [27], a comparison between trajectory-based methods and optimal control methods is provided.”
Learning perceptual coupling for motor primitives (2008, methodology): “The need for learning perceptual coupling in motor primitives has long been recognized in the motor primitive community [4].”
Imitation Learning of Positional and Force Skills Demonstrated via Kinesthetic Teaching and Haptic Input (2011, background): “[19] and further extended in Refs [20, 21] (see Ref.”
Dynamical Movement Primitives: Learning Attractor Models for Motor Behaviors (2013, background): “, 2009), reinforcement learning (Peters & Schaal, 2008; Kober & Peters, 2009), models for biological movement generation (Schaal et al., 2007; Hoffmann et al., 2009), and generating libraries of motor skills (Pastor et al.”
Movement templates for learning of hitting and batting (2010, methodology): “Since then, it has been simpliﬁed and in [3] it was shown that a single ﬁrst order system sufﬁces ˙ z = − τα z z. (1) This canonical system has the time constant τ = 1 /T where T is the duration of the motor primitive, a parameter α z which is chosen such that z ≈ 0 at T to ensure that the inﬂuence…”
Learning to select and generalize striking movements in robot table tennis (2012, background): “where τ is a temporal scaling factor and αz is a predefined constant which is chosen such that the behavior is stable (Ijspeert et al., 2002; Schaal et al., 2007).”
Imitation Learning (2017, background): “…with a dynamic system of motor primitives has produced a number of applications for learning discrete and rhythmic motions [Ijspeert et al. 2002a; Schaal et al. 2007; Kober et al. 2008], though most approaches focus on direct imitation without further optimization [Kober and Peters 2009a].”
Robot motor skill coordination with EM-based Reinforcement Learning (2010, background): “We extend here the use of DMP by considering synergy across the different motion variables through the association of a full matrix with each of the primitives (or states) instead of a fixed gain.”
Imitation and Reinforcement Learning (2010, methodology): “Since its inception in [2], the representation has been simplified, and it can be shown that a single, first-order system _ z 1⁄4 sazz suffices as a canonical system [11].”
Task-Specific Generalization of Discrete and Periodic Dynamic Movement Primitives (2010, methodology): “They are set so that the convergence of the underlying dynamic system is ensured [20].”