Sufficiently Accurate Model Learning
arXiv:1902.06862
Abstract
Modeling how a robot interacts with the environment around it is an important prerequisite for designing control and planning algorithms. In fact, the performance of controllers and planners is highly dependent on the quality of the model. One popular approach is to learn data driven models in order to compensate for inaccurate physical measurements and to adapt to systems that evolve over time. In this paper, we investigate a method to regularize model learning techniques to provide better error characteristics for traditional control and planning algorithms. This work proposes learning "Sufficiently Accurate" models of dynamics using a primal-dual method that can explicitly enforce constraints on the error in pre-defined parts of the state-space. The result of this method is that the error characteristics of the learned model is more predictable and can be better utilized by planning and control algorithms. The characteristics of Sufficiently Accurate models are analyzed through experiments on a simulated ball paddle system.
arXiv commentICRA 2020
References in corpus (12)
- Adam: A Method for Stochastic Optimization
- Model-Agnostic Meta-Learning for Fast Adaptation of Deep Networks
- DeepMind Control Suite
- Generative Adversarial Imitation Learning
- Visual Reinforcement Learning with Imagined Goals
- Learning Optimal Resource Allocations in Wireless Systems
- Differentiable MPC for End-to-end Planning and Control
- The Predictron: End-To-End Learning and Planning
- Constrained Policy Optimization
- Memory Augmented Control Networks
- SE3-Pose-Nets: Structured Deep Dynamics Models for Visuomotor Planning and Control
- GP-ILQG: Data-driven Robust Optimal Control for Uncertain Nonlinear Dynamical Systems