Minimax Differential Dynamic Programming: An Application to Robust Biped Walking

By Jun Morimoto, Christopher G. Atkeson. 2002

108 citations recorded by the source corpus.

How researchers see this paper

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

Citations in the corpus

Neural Network Dynamics for Model-Based Deep Reinforcement Learning with Model-Free Fine-Tuning (2017, background): “Model-based algorithms are known in general to outperform model-free learners in terms of sample complexity [4], and in practice have been applied successfully to control both simulated and real-world robotic systems, such as pendulums [6], legged robots [18], swimmers [19], and manipulators [20].”
Algorithmic Framework for Model-based Deep Reinforcement Learning with Theoretical Guarantees (2018, methodology): “…(Deisenroth et al., 2013) and has been successfully applied to robotics in both simulation and in the real world (Deisenroth & Rasmussen, 2011b; Morimoto & Atkeson, 2003; Deisenroth et al., 2011) using dynamical models ranging from Gaussian process (Deisenroth & Rasmussen, 2011b; Ko & Fox,…”
Neural probabilistic motor primitives for humanoid control (2018, methodology): “These can be obtained by sample-based optimization (e.g. Ding et al. (2015)) or by differential dynamic programming (Morimoto & Atkeson, 2003; Tassa et al., 2012; 2014).”
Robust Reinforcement Learning (2005, unknown): “We will apply RRL to more complex tasks like learning stand-up behaviors (Morimoto & Doya, 2000, 2001a) and biped locomotion (Atkeson & Morimoto, 2003; Morimoto & Atkeson, 2003 ) in future work.”
Probabilistic Differential Dynamic Programming (2014, methodology): “Various parametric and semi-parametric approaches have been developed to address these issues, such as minimax DDP using Receptive Field Weighted Regression (RFWR) by Morimoto and Atkeson [8], and DDP using expert-demonstrated trajectories by Abbeel et al. [9].”
Using inaccurate models in reinforcement learning (2006, background): “, Zhou et al. (1995); Dullerud and Paganini (2000) for more details on robust control.”
Learning variable impedance control (2011, methodology): “An example of a DDP application to robotics is presented in Morimoto and Atkeson (2002).”
Reinforcement learning of motor skills with policy gradients (2008, unknown): “with αn = [ε(∇J(θ)TF−1 θ ∇J(θ))−1]1/2, see Peters (2007) for derivations.”
Nonparametric Representation of Policies and Value Functions: A Trajectory-Based Approach (2002, methodology): “How we solve Isaacs’ equation and an application of this method are described in the companion paper [15].”
Receding Horizon Differential Dynamic Programming (2007, methodology): “In [6] a minimax variant of DDP is used to learn a controller for bipedal walking, again by designing a reference trajectory and rewarding the walker for tracking it.”
Information theoretic MPC for model-based reinforcement learning (2017, methodology): “Other techniques for model-based reinforcement learning incorporate trajectory optimization with model learning [9] or disturbance learning [10].”
Hierarchical Apprenticeship Learning with Application to Quadruped Locomotion (2007, background): “Prior work on legged locomotion has mostly focused on genera ting gaits for stably traversing fairly flat terrain (see, among many others, [9], [7]).”
Sample-based informationl-theoretic stochastic optimal control (2014, methodology): “Subsequently, the optimal controller is computed analytically on the linearised system and the resulting trajectory is used as a new linearisation point.”
A Survey on Policy Search for Robotics (2013, methodology): “RL approaches that explicitly address the problem of inaccurate models in robotics have only been introduced recently [1, 7, 21, 22, 35, 46, 52, 54, 73].”
Infinite-Horizon Model Predictive Control for Periodic Tasks with Contacts (2011, background): “Here we only repeat the main equations (as presented by [17]), and refer the interested reader to [12] for further details.”