检索结果-内蒙古大学图书馆

Data-Efficient and Safe learning for Humanoid Locomotion Aided by a Dynamic Balancing model

IEEE ROBOTICS AND AUTOMATION LETTERS 2020年第3期5卷 4376-4383页

作者： Ahn, Junhyeok Lee, Jaemin Sentis, Luis Univ Texas Austin Dept Mech Engn Austin TX 78712 USA Univ Texas Austin Dept Aerosp Engn & Engn Mech Austin TX 78712 USA

In this letter, we formulate a novel Markov Decision Process (MDP) for safe and data-efficient learning for humanoid locomotion aided by a dynamic balancing model. In our previous studies of biped locomotion, we relied on a low-dimensional robot model, commonly used in high-level Walking Pattern Generators (WPGs). However, a low-level feedback controller cannot precisely track desired footstep locations due to the discrepancies between the full order model and the simplified model. In this study, we propose mitigating this problem by complementing a WPG with reinforcement learning. More specifically, we propose a structured footstep control method consisting of a WPG, a neural network, and a safety controller. The WPG provides an analytical method that promotes efficient learning while the neural network maximizes long-term rewards, and the safety controller encourages safe exploration based on step capturability and the use of control-barrier functions. Our contributions include the following (1) a structured learning control method for locomotion, (2) a data-efficient and safe learning process to improve walking using a physics-based model, and (3) the scalability of the procedure to various types of humanoid robots and walking.

关键词： Humanoid and bipedal locomotion deep learning in robotics and automation model learning for control

来源：评论

学校读者我要写书评

暂无评论

Aggressive Perception-Aware Navigation Using Deep Optical Flow Dynamics and PixelMPC

引用

IEEE ROBOTICS AND AUTOMATION LETTERS 2020年第2期5卷 1207-1214页

作者： Lee, Keuntaek Gibson, Jason Theodorou, Evangelos A. Georgia Inst Technol Autonomous Control & Decis Syst Lab Atlanta GA 30332 USA

Recently, vision-based control has gained traction by leveraging the power of machine learning. In this work, we couple a model predictive control (MPC) framework to a visual pipeline. We introduce deep optical flow (DOF) dynamics, which is a combination of optical flow and robot dynamics. Using the DOF dynamics, MPC explicitly incorporates the predicted movement of relevant pixels into the planned trajectory of a robot. Our implementation of DOF is memory-efficient, data-efficient, and computationally cheap so that it can be computed in real-time for use in an MPC framework. The suggested Pixel model Predictive control (PixelMPC) algorithm controls the robot to accomplish a high-speed racing task while maintaining visibility of the important features (gates). This improves the reliability of vision-based estimators for localization and can eventually lead to safe autonomous flight. The proposed algorithm is tested in a photorealistic simulation with a high-speed drone racing task.

关键词： model learning for control optimization and optimal control visual servoing visual tracking visual-based navigation

来源：评论

学校读者我要写书评

暂无评论

q-VAE for Disentangled Representation learning and Latent Dynamical Systems

引用

IEEE ROBOTICS AND AUTOMATION LETTERS 2020年第4期5卷 5669-5676页

作者： Kobayashis, Taisuke Nara Inst Sci & Technol Div Informat Sci Nara 6300192 Japan

A variational autoencoder (VAE) derived from Tsallis statistics called q-VAE is proposed. In the proposed method, a standard VAE is employed to statistically extract latent space hidden in sampled data, and this latent space helps make robots controllable in feasible computational time and cost. To improve the usefulness of the latent space, this letter focuses on disentangled representation learning, e.g., beta-VAE, which is the baseline for it. Starting from a Tsallis statistics perspective, a new lower bound for the proposed q-VAE is derived to maximize the likelihood of the sampled data, which can be considered an adaptive beta-VAE with deformed Kullback-Leibler divergence. To verify the benefits of the proposed q-VAE, a benchmark task to extract the latent space from the MNIST dataset was performed. The results demonstrate that the proposed q-VAE improved disentangled representation while maintaining the reconstruction accuracy of the data. In addition, it relaxes the independency condition between data, which is demonstrated by learning the latent dynamics of nonlinear dynamical systems. By combining disentangled representation, the proposed q-VAE achieves stable and accurate long-term state prediction from the initial state and the action sequence.

关键词： Novel deep learning methods model learning for control

来源：评论

学校读者我要写书评

暂无评论

learning Dynamics for Improving control of Overactuated Flying Systems

引用

IEEE ROBOTICS AND AUTOMATION LETTERS 2020年第4期5卷 5283-5290页

作者： Zhang, Weixuan Brunner, Maximilian Ott, Lionel Kamel, Mina Siegwart, Roland Nieto, Juan Swiss Fed Inst Technol Autonomous Syst Lab Dept Mech & Proc Engn CH-8092 Zurich Switzerland Univ Sydney Camperdown NSW 2006 Australia

Overactuated omnidirectional flying vehicles are capable of generating force and torque in any direction, which is important for applications such as contact-based industrial inspection. This comes at the price of an increase in model complexity. These vehicles usually have non-negligible, repetitive dynamics that are hard to model, such as the aerodynamic interference between the propellers. This makes it difficult for high-performance trajectory tracking using a model-based controller. This letter presents an approach that combines a data-driven and a first-principle model for the system actuation and uses it to improve the controller. In a first step, the first-principle model errors are learned offline using a Gaussian Process (GP) regressor. At runtime, the first-principle model and the GP regressor are used jointly to obtain control commands. This is formulated as an optimization problem, which avoids ambiguous solutions present in a standard inverse model in overactuated systems, by only using forward models. The approach is validated using a tilt-arm overactuated omnidirectional flying vehicle performing attitude trajectory tracking. The results show that with our proposed method, the attitude trajectory error is reduced by 32% on average as compared to a nominal PID controller.

关键词： Aerial Systems Mechanics and control model learning for control

来源：评论

学校读者我要写书评

暂无评论

learning dynamical systems with bifurcations

引用

ROBOTICS AND AUTONOMOUS SYSTEMS 2021年 136卷 103700-103700页

作者： Khadivar, Farshad Lauzana, Ilaria Billard, Aude Ecole Polytech Fed Lausanne EPFL Learning Algorithms & Syst Lab LASA Route Cantonale CH-1015 Lausanne Switzerland

Trajectory planning through dynamical systems (DS) provides robust control for robots and has found numerous applications from locomotion to manipulation. However, to date, DS for controlling rhythmic patterns are distinct from DS used to control point to point motion and current approaches switch at run time across these to enable multiple behaviors. This switching can be brittle and subject to instabilities. We present an approach to embed cyclic and point to point dynamics in a single DS. We offer a method to learn the parameters of complete DS through a two-step optimization. By exploiting Hopf bifurcations, we can explicitly and smoothly transit across periodic and non-periodic phases, linear and nonlinear limit cycles, and non-periodic phases, in addition to changing the equilibrium's location and the limit cycle's amplitude. We use diffeomorphism and learn a mapping to modify the learned limit cycle to generate nonlinear limit cycles. The approach is validated with a real 7 DOF KUKA LWR 4+ manipulator to control wiping and with a humanoid robot in simulation. (C) 2020 The Author(s). Published by Elsevier B.V.

关键词： learning from demonstration model learning for control Motion control Optimization and optimal control

来源：评论

学校读者我要写书评

暂无评论

Multi-objective symbolic regression for physics-aware dynamic modeling

引用

EXPERT SYSTEMS WITH APPLICATIONS 2021年 182卷 1页

作者： Kubalik, Jiri Derner, Erik Babuska, Robert Czech Tech Univ Czech Inst Informat Robot & Cybernet Prague 16000 Czech Republic Czech Tech Univ Fac Elect Engn Dept Control Engn Prague 16627 Czech Republic Delft Univ Technol Cognit Robot NL-2628 CD Delft Netherlands

Virtually all dynamic system control methods benefit from the availability of an accurate mathematical model of the system. This includes also methods like reinforcement learning, which can be vastly sped up and made safer by using a dynamic system model. However, obtaining a sufficient amount of informative data for constructing dynamic models can be difficult. Consequently, standard data-driven model learning techniques using small data sets that do not cover all important properties of the system yield models that are partly incorrect, for instance, in terms of their steady-state characteristics or local behavior. However, often some knowledge about the desired physical properties of the model is available. Recently, several symbolic regression approaches making use of such knowledge to compensate for data insufficiency were proposed. Therefore, this knowledge should be incorporated into the model learning process to compensate for data insufficiency. In this paper, we consider a multi-objective symbolic regression method that optimizes models with respect to their training error and the measure of how well they comply with the desired physical properties. We propose an extension to the existing algorithm that helps generate a diverse set of high-quality models. Further, we propose a method for selecting a single final model out of the pool of candidate output models. We experimentally demonstrate the approach on three real systems: the TurtleBot 2 mobile robot, the Parrot Bebop 2 drone and the magnetic manipulation system. The results show that the proposed model-learning algorithm yields accurate models that are physically justified. The improvement in terms of the model's compliance with prior knowledge over the models obtained when no prior knowledge was involved in the learning process is of several orders of magnitude.

关键词： Symbolic regression Genetic programming model learning for control Motion control

来源：评论

学校读者我要写书评

暂无评论

Provably Robust learning-Based Approach for High-Accuracy Tracking control of Lagrangian Systems

引用

IEEE ROBOTICS AND AUTOMATION LETTERS 2019年第2期4卷 1587-1594页

作者： Helwa, Mohamed K. Heins, Adam Schoellig, Angela P. Univ Toronto Inst Aerosp Studies Dynam Syst Lab Toronto ON M3H 5T6 Canada Cairo Univ Dept Elect Power & Machines Giza 12613 Egypt

Lagrangian systems represent a wide range of robotic systems, including manipulators, wheeled and legged robots, and quadrotors. Inverse dynamics control and feedforward linearization are typically used to convert the complex nonlinear dynamics of Lagrangian systems to a set of decoupled double integrators, and then a standard, outer-loop controller can be used to calculate the commanded acceleration for the linearized system. However, these methods typically depend on having a very accurate system model, which is often not available in practice. While this challenge has been addressed in the literature using different learning approaches, most of these approaches do not provide safety guarantees in terms of stability of the learning-based control system. In this letter, we provide a novel, learning-based control approach based on Gaussian processes (GPs) that ensures both stability of the closed-loop system and high-accuracy tracking. We use GPs to approximate the error between the commanded and the actual acceleration of the system, and then use the predicted mean and variance of the GP to calculate an upper bound on the uncertainty of the linearized model. This uncertainty bound is then used in a robust, outer-loop controller to ensure stability of the overall system. Moreover, we show that the tracking error converges to a ball with a radius that can be made arbitrarily small. Finally, we verify the effectiveness of our approach via simulations on a 2 degree-of-freedom (DOF) planar manipulator and experimentally on a 6 DOF industrial manipulator.

关键词： model learning for control robot safety robust/adaptive control of robotic systems learning and adaptive systems

来源：评论

学校读者我要写书评

暂无评论

Learn Fast, Forget Slow: Safe Predictive learning control for Systems With Unknown and Changing Dynamics Performing Repetitive Tasks

引用

IEEE ROBOTICS AND AUTOMATION LETTERS 2019年第2期4卷 2180-2187页

作者： McKinnon, Christopher D. Schoellig, Angela P. Univ Toronto Inst Aerosp Studies Dynam Syst Lab N York ON M3H 5T6 Canada

We present a control method for improved repetitive path following for a ground vehicle that is geared toward long-term operation, where the operating conditions can change over time and are initially unknown. We use weighted Bayesian linear regression (wBLR) to model the unknown dynamics, and show how this simple model is more accurate in both its estimate of the mean behavior and model uncertainty than Gaussian process regression and generalizes to novel operating conditions with little or no tuning. In addition, wBLR allows us to use fast adaptation and long-term learning in one unified framework to adapt quickly to new operating conditions and learn repetitive model errors over time. This comes with the added benefit of lower computational cost, longer look-ahead, and easier optimization when the model is used in a stochastic model-predictive controller (MPC). In order to fully capitalize on the long prediction horizons that are possible with this new approach, we use Tube MPC to reduce the growth of predicted uncertainty. We demonstrate the effectiveness of our approach in the experiment on a 900-kg ground robot showing results over 3.0 km of driving with both physical and artificial changes to the robot's dynamics. All of our experiments are conducted using a stereo camera for localization.

关键词： learning and adaptive systems model learning for control robot safety field robots

来源：评论

学校读者我要写书评

暂无评论

Autonomous Precision Pouring From Unknown Containers

引用

IEEE ROBOTICS AND AUTOMATION LETTERS 2019年第3期4卷 2317-2324页

作者： Kennedy, Monroe Schmeckpeper, Karl Thakur, Dinesh Jiang, Chenfanfu Kumar, Vijay Daniilidis, Kostas Univ Penn Grasp Lab Philadelphia PA 19104 USA Univ Penn SIG Ctr Comp Graph Philadelphia PA 19104 USA

We autonomously pour from unknown symmetric containers found in a typical wet laboratory for the development of a robot-assisted, rapid experiment preparation system. The robot estimates the pouring container symmetric geometry, then leverages simulated pours as priors for a given fluid to pour precisely and quickly in a single attempt. The fluid is detected in the transparent receiving container by combining weight and vision. The change of volume in the receiver is a function of the geometry of the pouring container, the pouring angle, and rate. To determine the volumetric flow rate, the profile for maximum containable volume for a given angle is estimated along with the time delay of the fluid exiting the container. A trapezoidal trajectory generation algorithm prescribes the desired volumetric flow rate as a function of the estimation accuracy. A hybrid control strategy is then used to attenuate volumetric error. Three methods are compared for estimating the volume-angle profile, and it is shown that a combination of online system identification and leveraged model priors results in reliable performance. The major contributions of this work are a system capable of pouring quickly and precisely from varying symmetric containers in a single attempt with limited priors, and a novel fluid detection method. This system is implemented on the Rethink Robotics Sawyer and KUKA LBR iiwa manipulators.

关键词： model learning for control motion control manipulation planning service robots

来源：评论

学校读者我要写书评

暂无评论

Robot Cooperative Behavior learning Using Single-Shot learning From Demonstration and Parallel Hidden Markov models

引用

IEEE ROBOTICS AND AUTOMATION LETTERS 2019年第2期4卷 193-200页

作者： Lafleche, Jean-Francois Saunderson, Shane Nejat, Goldie Univ Toronto Dept Mech & Ind Engn Autonomous Syst & Biomechatron Lab Toronto ON M6S3G8 Canada

For robots to become collaborative assistants, they need to be capable of naturally interacting with users in real environments. They also need to be able to learn new skills from non-expert users. In this letter, we present a novel parallel hidden Markov model (PaHMM) architecture for learning from demonstration (LfD), which allows a robot to learn a sequence of cooperative and non-cooperative behaviors from a single demonstration (single-shot) of a task-based human-robot interaction from two interacting teachers. During teaching, the robot learns both a human-robot interactionmodel and an object interactionmodel in order to be able to effectively determine its own behaviors. Experiments with a Baxter robot and several teachers were conducted to validate the ability of the robot to learn both cooperative and non-cooperative behaviors during a task-based interaction. Comparison experiments also show the robustness of our approach to spatial variations from the demonstrated behaviors and tracking errors when compared to other approaches.

关键词： learning from demonstration (LfD) physical human-robot interaction model learning for control

来源：评论

学校读者我要写书评

暂无评论

建议与咨询留下您的常用邮箱和电话号码，以便我们向您反馈解决方案和替代方法

时间限定

文献类型

馆藏选择

核心期刊

语言

文献类型

帮助

文字说明：

检索规则说明：

检索范例：

分类表

所选分类

限定检索结果

文献类型

馆藏范围

日期分布

学科分类号

主题

机构

作者

语言

请选择保存的检索档案：

请选择收藏分类：

通借通还

建议与咨询 留下您的常用邮箱和电话号码，以便我们向您反馈解决方案和替代方法

时间限定

文献类型

馆藏选择

核心期刊

语言

文献类型

帮助

文字说明：

检索规则说明：

检索范例：

分类表

所选分类

限定检索结果

文献类型

馆藏范围

日期分布

学科分类号

主题

机构

作者

语言

请选择保存的检索档案： 新增检索档案 确定 取消

请选择收藏分类： 新增自定义分类 确定 取消

通借通还

建议与咨询留下您的常用邮箱和电话号码，以便我们向您反馈解决方案和替代方法

请选择保存的检索档案：

请选择收藏分类：