检索结果-内蒙古大学图书馆

Event-Triggered learning-Based control of Quadrotors for Accurate Agile Trajectory Tracking

IEEE ROBOTICS AND AUTOMATION LETTERS 2024年第6期9卷 5855-5862页

作者： Zhang, Chenxu Li, Xiaohan Wang, Xinghu Ji, Haibo Univ Sci & Technol China Dept Automat Hefei 230027 Peoples R China Univ Sci & Technol China Anhui Prov Key Lab Intelligent Low Carbon Informat Hefei Peoples R China

Accurate tracking control of agile quadrotors is challenging because of complicated aerodynamic effects. In this letter, we develop a tunable event-triggered learning-based control framework for quadrotors to achieve agile trajectory tracking in the presence of aerodynamic effects. A novel Gaussian process (GP)-augmented differential flatness-based controller is proposed to produce collective thrusts and body rates (CTBR) control commands to perform agile trajectory tracking. To achieve a better trade-off between the computational burden of GP learning and the tracking accuracy, a tunable event-triggered learning strategy is designed by using the Lyapunov redesign technique. We verify that the developed framework can effectively improve the performance of agile trajectory tracking of quadrotors and release the computational burden of GP learning in the presence of aerodynamic effects through both simulations and real-world experiments. Experimental results show that our framework can reduce the computational burden by 78.4% while maintaining satisfiable tracking accuracy.

关键词： Aerial systems: mechanics and control model learning for control agile flight Gaussian process

来源：评论

学校读者我要写书评

暂无评论

Magnetic manipulation of unknown and complex conductive nonmagnetic objects with application in the remediation of space debris

引用

INTERNATIONAL JOURNAL OF ROBOTICS RESEARCH 2024年第7期44卷 1117-1137页

作者： Tabor, Griffin F. Pham, Lan N. Abbott, Jake J. Hermans, Tucker Univ Utah Robot Ctr Salt Lake City UT 84112 USA Univ Utah Kahlert Sch Comp Salt Lake City UT 84112 USA Univ Utah Dept Mech Engn Salt Lake City UT 84112 USA NVIDIA Seattle WA USA

This article extends recent work in magnetic manipulation of conductive, nonmagnetic objects using rotating magnetic dipole fields. Eddy-current-based manipulation provides a contact-free way to manipulate metallic objects. We are particularly motivated by the large amount of aluminum in space debris. We previously demonstrated dexterous manipulation of solid spheres with all object parameters known a priori. This work expands the previous model, which contained three discrete modes, to a continuous model that covers all possible relative positions of the manipulated spherical object with respect to the magnetic field source. We further leverage this new model to examine manipulation of spherical objects with unknown physical parameters by applying techniques from the online-optimization and adaptive-control literature. Our experimental results validate our new dynamics model, showing that we get improved performance compared to the previously proposed model, while also solving a simpler optimization problem for control. We further demonstrate the first physical magnetic manipulation of aluminum spheres, as previous controllers were only physically validated on copper spheres. We show that our adaptive control framework can quickly acquire useful object parameters when weakly initialized. Finally, we demonstrate that the spherical-object model can be used as an approximate model for adaptive control of nonspherical objects by performing magnetic manipulation of a variety of objects for which a spherical model is not an obvious approximation.

关键词： Space robotics and automation dexterous manipulation model learning for control robust/adaptive control optimization and optimal control

来源：评论

学校读者我要写书评

暂无评论

Accurate Robotic Pushing Manipulation Through Online model Estimation Under Uncertain Object Properties

引用

IEEE ROBOTICS AND AUTOMATION LETTERS 2024年第10期9卷 8730-8737页

作者： Lee, Yongseok Kim, Keehoon Pohang Univ Sci & Technol POSTECH Dept Mech Engn Pohang 37673 South Korea

Robotic pushing is a fundamental non-prehensile manipulation skill essential for handling objects that are difficult to grasp. This letter proposes a highly accurate robotic pushing framework that utilizes an online estimated model to push objects along a given nominal trajectory, despite uncertain object properties such as friction coefficients, mass distribution, and the position of the center of friction (CoF). The core concept involves estimating an optimal pushing motion model capable of representing observed local motions. A generalized form of the conventional analytical model, coupled with a moving-window Unscented Kalman Filter (UKF), serves as the online estimated model. It captures the local behavior of the pushed objects and is integrated with a model predictive control-based pushing strategy to achieve precise pushing performance. In experiments, the proposed robotic pushing framework demonstrated superior accuracy in tracking the given nominal trajectory compared to the conventional analytical model and data-driven model approaches, even when the motion model was perturbed. Additionally, the practicality of the proposed framework was showcased through a demonstration involving an autonomous robot collecting dishes, illustrating its applicability in various real-world applications.

关键词： Manipulation planning model learning for control non-prehensile manipulation robotic planar pushing Manipulation planning model learning for control non-prehensile manipulation robotic planar pushing

来源：评论

学校读者我要写书评

暂无评论

Deep Dynamics: Vehicle Dynamics modeling With a Physics-Constrained Neural Network for Autonomous Racing

引用

IEEE ROBOTICS AND AUTOMATION LETTERS 2024年第6期9卷 5292-5297页

作者： Chrosniak, John Ning, Jingyun Behl, Madhur Univ Virginia Dept Comp Sci Charlottesville VA 22903 USA

Autonomous racing is a critical research area for autonomous driving, presenting significant challenges in vehicle dynamics modeling, such as balancing model precision and computational efficiency at high speeds (>280 km/h), where minor errors in modeling have severe consequences. Existing physics-based models for vehicle dynamics require elaborate testing setups and tuning, which are hard to implement, time-intensive, and cost-prohibitive. Conversely, purely data-driven approaches do not generalize well and cannot adequately ensure physical constraints on predictions. This letter introduces Deep Dynamics, a physics-constrained neural network (PCNN) for autonomous racecar vehicle dynamics modeling. It merges physics coefficient estimation and dynamical equations to accurately predict vehicle states at high speeds. A unique Physics Guard layer ensures internal coefficient estimates remain within their nominal physical ranges. Open-loop and closed-loop performance assessments, using a physics-based simulator and full-scale autonomous Indy racecar data, highlight Deep Dynamics as a promising approach for modeling racecar vehicle dynamics.

关键词： Deep learning methods model learning for control dynamics

来源：评论

学校读者我要写书评

暂无评论

Fine Robotic Manipulation Without Force/Torque Sensor

引用

IEEE ROBOTICS AND AUTOMATION LETTERS 2024年第2期9卷 1206-1213页

作者： Shan, Shilin Pham, Quang-Cuong Nanyang Technol Univ Sch Mech & Aerosp Engn Singapore 639798 Singapore Nanyang Technol Univ Eureka Robot Singapore 639798 Singapore Nanyang Technol Univ Singapore Ctr 3D Printing SC3DP Sch Mech & Aerosp Engn Singapore 639798 Singapore

Force Sensing and Force control are essential to many industrial applications. Typically, a 6-axis force/torque (F/T) sensor is installed between the robot's wrist and the end effector to measure the forces and torques exerted by the environment on the robot (the external wrench). While a typical 6-axis F/T sensor can offer highly accurate measurements, it is expensive and vulnerable to drift and external impacts. Existing methods aiming at estimating the external wrench using only the robot's internal signals are limited in scope. For instance, the estimation accuracy has mainly been validated in free-space motions and simple contacts, rather than tasks like assembly that require high-precision force control. In this letter, we present a Neural-Network-based solution to overcome these challenges. We offer a detailed discussion on model structure, training data categorization and collection, as well as fine-tuning strategies. These steps enable precise and reliable wrench estimations across a variety of scenarios. As an illustration, we demonstrate a pin insertion experiment with a 100-micron clearance and a hand-guiding experiment, both performed without external F/T sensors or joint torque sensors.

关键词： Dynamics industrial robots machine learning for robot control model learning for control

来源：评论

学校读者我要写书评

暂无评论

Real-Time Constrained Tracking control of Redundant Manipulators Using a Koopman-Zeroing Neural Network Framework

引用

IEEE ROBOTICS AND AUTOMATION LETTERS 2024年第2期9卷 1732-1739页

作者： Sah, Chandan Kumar Singh, Rajpal Keshavan, Jishnu Indian Inst Sci Dept Mech Engn Bangalore 560012 India

This letter proposes a combined Koopman-ZNN (Zeroing Neural Network) architecture for real-time control of redundant manipulators subject to input constraints. An autoencoder-based neural architecture is employed to discover the bilinear Koopman model for manipulator dynamics in joint space using input-output data, which is subsequently integrated with a feed-forward neural network that maps the joint coordinates to end-effector Cartesian coordinates. The proposed architecture allows for efficient learning of highly accurate models using a significantly lower number of observable states compared to the previous studies. This learning architecture is then coupled with a ZNN controller, which offers a computationally inexpensive alternative to state-of-the-art Nonlinear model Predictive control (NMPC) controllers whose computational burden might render real-time control infeasible. The low-dimensional nature of the learned model, combined with a computationally inexpensive ZNN controller, facilitates real-time control applications with improved tracking accuracy. Simulation and experimental studies of trajectory tracking, including performance comparisons with leading alternative designs, are used to verify the efficacy of the proposed scheme.

关键词： Redundant robots model learning for control machine learning for robot control

来源：评论

学校读者我要写书评

暂无评论

Identification and learning-Based control of an End-Effector for Targeted Throwing

引用

IEEE ROBOTICS AND AUTOMATION LETTERS 2024年第11期9卷 9558-9564页

作者： Pasala, Haasith Venkata Sai Govindan, Nagamanikandan Brahmbhatt, Samarth Int Inst Informat Technol Robot Res Ctr Hyderabad 500032 India Intel Lab Santa Clara CA 95054 USA

Drones and mobile manipulators that can throw objects in addition to picking and placing them can be highly useful in industrial automation, warehouse environments, search and rescue operations, and sports training. However, predominant end-effectors primarily cater to grasping functions, neglecting the throwing aspect. Currently, throwing is achieved by fast whole-arm motion (Zeng et al., 2020), an approach that raises concerns regarding safety and energy efficiency. Additionally, targeted throwing poses several challenges due to the uncertainties in model parameters and unmodelled dynamics. This letter presents a new end-effector mechanism that can grasp and then place or throw an object using stored elastic energy. The instantaneous release of this stored energy propels the grasped object into projectile motion, facilitating its placement in a desired target location which can lie beyond the reachable workspace of the robot arm. We describe the mechanical design of the end-effector, its simulation model, a system identification method to fit model parameters, and a data-driven residual learning framework. The residual model predicts control input residuals arising from model uncertainties, improving targeted throwing accuracy even with unseen objects. Experiments conducted with our robot arm mounted end-effector show the efficacy of our end-effector mechanism and associated algorithms for targeted throwing.

关键词： Robots End effectors Motors Grasping Gears Solid modeling Service robots Grippers and other end-effectors throwing manipulation model learning for control mechanism design and system identification

来源：评论

学校读者我要写书评

暂无评论

Multi-Tap Resistive Sensing and FEM modeling Enables Shape and Force Estimation in Soft Robots

引用

IEEE ROBOTICS AND AUTOMATION LETTERS 2024年第3期9卷 2830-2837页

作者： Tian, Sizhe Cangan, Barnabas Gavin Navarro, Stefan Escaida Beger, Artem Duriez, Christian Katzschmann, Robert K. Swiss Fed Inst Technol Soft Robot Lab IRIS D MAVT CH-8092 Zurich Switzerland Univ O Higgins Inst Ciencias Ingn Rancagua 2820000 Chile Lille Univ INRIA Lille Nord Europe F-59000 Lille France Festo SE & Co KG D-73734 Esslingen Germany

We address the challenge of reliable and accurate proprioception in soft robots, specifically soft robots with tight packaging constraints and relying only on internally embedded sensors. While various sensing approaches with single sensors have commonly been tried using a constant curvature assumption, we look into sensing local deformations at multiple sensor locations. In our approach, we multi-tap an off-the-shelf resistive sensor by creating multiple electrical connections onto the resistive layer of the sensor, and we insert the sensor into a soft body. This modification allows us to measure changes in resistance at multiple segments throughout the length of the sensor, providing improved resolution of local deformations in the soft body. These measurements inform a model based on a finite element method (FEM) that estimates the shape of the soft body and the magnitude of an external force acting at a known arbitrary location. Our model-based approach estimates soft body deformation with approximately 3% average relative error while taking into account internal fluidic actuation. Our estimate of external force disturbance has an 11% relative error within a range of 0 to 5 N. For instance, the combined sensing and modeling approach can be integrated into soft manipulation platforms to enable features such as identifying the shape and material properties of an object being grasped. Such manipulators can benefit from the inherent softness and compliance while being fully proprioceptive, relying only on embedded sensing and not on external systems such as motion capture. Such proprioception is essential for the deployment of soft robots in real-world scenarios.

关键词： Robot sensing systems Shape Soft robotics Sensors Flexible printed circuits Force Estimation Flexible robotics force control force and tactile sensing model learning for control soft robot applications soft robot materials and design soft sensors and actuators

来源：评论

学校读者我要写书评

暂无评论

Incremental Sparse Gaussian Process-Based model Predictive control for Trajectory Tracking of Unmanned Underwater Vehicles

引用

IEEE ROBOTICS AND AUTOMATION LETTERS 2025年第3期10卷 2327-2334页

作者： Dang, Yukun Huang, Yao Shen, Xuyu Zhu, Daqi Chu, Zhenzhong Univ Shanghai Sci & Technol Dept Mech Engn Shanghai 200093 Peoples R China

In this letter, a model Predictive control (MPC) approach based on the Incremental Sparse Gaussian Process (ISGP) is designed for trajectory tracking of Unmanned Underwater Vehicles (UUVs). The performance of MPC depends on the accuracy of system modeling. However, building an accurate dynamic model for the UUV is challenging due to imprecise hydrodynamic coefficients and strong nonlinearities. Thus, the Gaussian Process (GP) is employed to regress the deviating parts of the system model. A sparsification rule is proposed to reduce the training dataset size by removing less valuable data, thereby simplifying the complexity of GP regression training. Additionally, a method for incrementally updating the training data is provided, along with a rigorous stability proof. Finally, simulations are conducted in a third-party ROS environment to demonstrate the efficiency and accuracy of the proposed method.

关键词： Mathematical models Training Kernel Uncertainty Trajectory Trajectory tracking control systems Kinematics Gaussian processes Adaptation models Marine robotics model learning for control motion control

来源：评论

学校读者我要写书评

暂无评论

Freeway Traffic modeling by Physics-Regularized Gaussian Processes

IEEE OPEN JOURNAL OF INTELLIGENT TRANSPORTATION SYSTEMS

引用

IEEE OPEN JOURNAL OF INTELLIGENT TRANSPORTATION SYSTEMS 2025年 6卷 116-130页

作者： Binjaku, Kleona Pasquale, C. Mece, E. K. Sacone, S. Univ Genoa Dept Informat Bioengn Robot & Syst Engn I-16131 Genoa Italy Polytech Univ Tirana Dept Comp Engn Tirana 1000 Albania

Effective traffic management and control are essential for mitigating congestion and minimizing environmental impacts on road transportation systems. In this paper, we propose a novel approach for traffic modeling that integrates physics-based dynamics with machine learning techniques. Our method leverages Gaussian Processes (GPs) and a multi-class second-order discrete traffic model known as METANET to develop a Physics-Regularized Machine learning framework. Furthermore, the proposed approach includes for the first time multi-class on/off ramps within the modeling framework, enhancing the realism of the predictive model. We systematically evaluate the performance of the hybrid model across varying dataset sizes to determine optimal data requirements for accurate traffic predictions. Experimental results indicate the improved predictive performance of the proposed approach compared to traditional machine learning and physics-based models. Our findings underscore the potential of Physics-Regularized Machine learning for enhancing traffic management and control strategies in real-world scenarios.

关键词： Computational modeling Predictive models Data models Machine learning Neural networks Traffic control Adaptation models Real-time systems Gaussian processes Accuracy Intelligent transportation systems machine learning 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：