A5085578333- Robot Manipulation and Learning
- Teleoperation and Haptic Systems
- Adaptive Control of Nonlinear Systems
- Adaptive Dynamic Programming Control
- Iterative Learning Control Systems
- Hand Gesture Recognition Systems
- Elevator Systems and Control
- Distributed Control Multi-Agent Systems
- Gaze Tracking and Assistive Technology
- Reinforcement Learning in Robotics
- Soft Robotics and Applications
- Mechanical Circulatory Support Devices
- Ionosphere and magnetosphere dynamics
- Advanced Vision and Imaging
- Machine Learning in Healthcare
- Photovoltaic Systems and Sustainability
- Machine Learning and ELM
- Advanced Wireless Network Optimization
- Piezoelectric Actuators and Control
- Earthquake Detection and Analysis
- COVID-19 diagnosis using AI
- Muscle activation and electromyography studies
- Global Energy and Sustainability Research
- Advanced Adaptive Filtering Techniques
- Hybrid Renewable Energy Systems
Nanjing University of Information Science and Technology
2022-2024
IRD Fuel Cells (Denmark)
2024
University of Macau
2019-2021
South China University of Technology
2016-2018
In this paper, an admittance adaptation method has been developed for robots to interact with unknown environments. The environment be interacted is modeled as a linear system. the presence of dynamics environments, observer in robot joint space employed estimate interaction torque, and control adopted regulate behavior at points. An adaptive neural controller using radial basis function guarantee trajectory tracking. A cost that defines performance torque regulation tracking minimized by...
In this paper, we present a sensorless admittance control scheme for robotic manipulators to interact with unknown environments in the presence of actuator saturation. The external environment is defined as linear models dynamics. Using control, manipulator controlled be compliant torque from environment. acted on end-effector estimated by using disturbance observer based generalized momentum. model uncertainties are solved radial basis neural networks (NNs). To guarantee tracking...
In this paper, a force sensorless control scheme based on neural networks (NNs) is developed for interaction between robot manipulators and human arms in physical collision. scheme, the trajectory generated by using geometry vector method with Kinect sensor. To comply external torque from environment, paper presents admittance approach joint space an observer approach, which used to estimate torques applied operator. deal tracking problem of uncertain manipulator, adaptive controller...
In this paper, an adaptive admittance control scheme is developed for robots to interact with time-varying environments. Admittance adopted achieve a compliant physical robot–environment interaction, and the uncertain environment dynamics defined as linear system. A critic learning method used obtain desired parameters based on cost function composed of interaction force trajectory tracking without knowledge environmental dynamics. To deal dynamic uncertainties in system, neural-network...
This article presents a neural network based admittance control scheme for robotic manipulators when interacting with the unknown environment in presence of actuator deadzone without needing force sensing. A compliant behavior response to external torques from is achieved by control. Inspired broad learning system, flatted structure using radial basis function (RBF) incremental algorithm proposed estimate torque, which can avoid retraining process if system modeled insufficiently. To deal...
In this article, a robust control scheme is proposed for robots to achieve an optimal performance in the process of interacting with external forces from environments. The environmental dynamics are defined as linear model, and interaction evaluated by cost function, which composed trajectory errors force regulation. Based on admittance control, reference adaptation method used minimize function performance. To make tracking controller unknown disturbance internal system dynamics, auxiliary...
This article proposes a spatial learning control system for robots to achieve desired behavior during interacting with unknown environments. In contacting the environment, force is estimated by observer, so sensing devices are not required. Motivated human interaction versatility, reference trajectory of robot updating law such that can be maintained at level. Compared iteration algorithm based on time domain, which requires maintaining fixed motion speed each iteration, proposed method...
Abstract In this study, the Empirical Mode Decomposition algorithm (EMD) and Long Short Term Memory neural network (LSTM) are combined into an EMD‐LSTM model, to predict variation of >2 MeV electron fluxes 1 day ahead. Input parameters include Pc5 power, AP, AE, Kp, >0.6 MeV, historical flux values, used for predictions. All time resolution daily integral values. As compared prediction results model with other classical models, show that ahead efficiency possesses a 0.80, highest can...
In this article, a proactive control strategy is developed for robots interacting with humans by integrating the estimation of human partner's motion intention. A model used and least square-based observer employed to estimate input without force sensor. Using intention, neural network (NN)-enhanced robot controller designed make actively follow trajectory. NNs are integrated into approximate compensate system uncertainties, so that tracking performance can be guaranteed. Rigorous analysis...
Endowing robots with the ability to maintain precise interaction force is critical for performing control tasks in dynamic environments characterized by unknown and varying stiffness geometry, such as aircraft wing skins other thin, soft materials. This article presents an adaptive force-tracking admittance controller (AFTAC), ensuring tracking performance through meticulous design of both force-based outer loop position-based inner loop. First, a finite-time based on neural networks (NNs)...
The traditional 3D human motion capture methods require operators to wear tracking sensors on their body. However, these devices would bring much inconvenience. Currently, vision based systems provide us an alternative approach for technology. In this paper, kinematics and function approximation technique (FAT), a novel method is presented the trajectory control of Baxter robot. Each arm robot has seven degrees freedom. geometry vector applied by using Microsoft Kinect sensor. A FAT system...
In this paper, a sensorless admittance control scheme is developed for robot manipulators in the presence of input saturation by employing neural networks. To deal with system uncertainties, radial basis function network (RBFNN) integrated into design. order to saturation, compensator applied handle problem. interact environment, employed and external torque estimated using generalized momentum based disturbance observer. Simulations are performed verify effectiveness proposed scheme.
The COVID-19 virus has been raging around the world for months and killing more than a million people. It is extremely infectious due to its easy transmission long incubation period. Until now, number of people diagnosed with infection increasing dramatically each day. At this stage, fast, accurate, early clinical assessment disease severity vital. For purpose, machine learning tool an effective way diagnose it. To support decision-making logistical planning in healthcare systems, inspired...
A spatial iterative control method is developed for robots to interact with an unknown environment at a desired level. Motivated by the human adaptive behaviour, robot controller can adapt its reference trajectory maintain interaction force designing learning law. Considering uncertain dynamics of robot, algorithm integrating neural networks employed enable track trajectory, so that performance achieved. Through Lyapunov theory, signals closed-loop system are analyzed and proven be...
The confusion of concepts has been present in the emerging propositions energy sector. In research, we sort through new energy, green clean recyclable recycled and renewable order to clarify terms basic scientific understandings context primary (PE) production. We further categorize PE trends by their properties, i.e., sources from phosphates, geo-oscillation, biosynthesis, so as evaluate strengths weaknesses
This paper considers the study scenario that end-effector of a manipulator follows desired trajectory and interacts with external environment. To maximize interaction performance, admittance control is combined adaptive dynamic programming (ADP). The optimal parameters can be learned online without prior knowledge A data-driven Hybrid Iteration employed in ADP, which relax initial stabilizing requirement at same time has faster convergence rate compared Value Iteration. In addition, more...
Increasing the linear modulation range
Hand gesture recognition is highly valued for its potential applications in contactless human-computer interaction (HCI). Aiming at the problem that system based on ordinary camera susceptible to different lighting conditions and complex background environment, an improved algorithm depth image fingertip detection proposed. Firstly, obtained by zed stereo camera. The position of hand detected segmented information color information, It proposed combine convex hull k-curvature detect number...
Compliant motion adaptation is essential in the task of robot interacting with environment. This paper presents a planning framework that generates and tracks trajectory to achieve compliant behavior during interaction. First, motions from multiple demonstrations are modelled by dynamical system (DS) Gaussian Mixture Models (GMM), parameters DS model learning online. In case contacting force environment, admittance control employed modify generate motion. The estimated sensorless approach....
In this paper, we develop a learning controller that adapts and tracks the impedance trajectory for robots interacting with unknown environments. Impedance adaptation is used to compensate contacting environment, while reference maintain prescribed interaction force. The tracking performance ensured by an adaptive Integral Reinforcement (IRL) partially system dynamics. contact dynamics are analysed via Lyapunov theory effectiveness of proposed control method verified through simulations.