A5101816136- Prosthetics and Rehabilitation Robotics
- Robot Manipulation and Learning
- Adaptive Control of Nonlinear Systems
- Adaptive Dynamic Programming Control
- Muscle activation and electromyography studies
- Reinforcement Learning in Robotics
- Bayesian Methods and Mixture Models
- Advanced Mathematical Modeling in Engineering
- Teleoperation and Haptic Systems
- Iterative Learning Control Systems
- Distributed Control Multi-Agent Systems
- Modular Robots and Swarm Intelligence
Changchun University of Technology
2020-2025
Major challenges of controlling human-robot collaboration (HRC)-oriented modular robot manipulators (MRMs) include the estimation human motion intention while cooperating with a and performance optimization. This article proposes cooperative game-based approximate optimal control method MRMs for HRC tasks. A harmonic drive compliance model-based is developed using position measurements only, which forms basis MRM dynamic model. Based on differential game strategy, problem HRC-oriented...
To address the force/position control challenges in transitioning from free-space motion to tasks involving environmental contact, this paper proposes an Adaptive Dynamic Programming (ADP)-based finite-time optimal backstepping method for Reconfigurable Robot Manipulators (RRMs), which ensures rapid convergence of state errors under external constraints while maintaining system stability. By integrating robust control, proposed enhances both speed and robustness against uncertainties....
Bayesian inference for Dirichlet-Multinomial (DM) models has a long and important history. The concentration parameter $\alpha$ is pivotal in smoothing category probabilities within the multinomial distribution crucial afterward. Due to lack of tractable form its marginal likelihood, often chosen ad-hoc, or estimated using approximation algorithms. A constant leads inadequate probabilities, particularly sparse compositional count datasets. In this paper, we introduce novel class prior...
Abstract In this paper, a human motion intention estimation-based decentralized robust interaction control method of modular robot manipulators (MRMs) is proposed under the situation physical human–robot (pHRI). Different from traditional scheme that depends on biological signal and centralized method, implemented using only position measurements each joint module in investigation. Based harmonic drive compliance model, novel torque-sensorless estimation developed, which utilizes information...
In this paper, a decentralized robust control method of modular robot manipulator (MRM) is proposed under the situation physical human-robot interaction (pHRI). Different from traditional that relies on precise dynamic model and centralized scheme, based human motion intention, which estimated through radial basis function neural network (RBFNN). By only using local information each joint, controller designed to deal with various uncertainties joint trajectory tracking problems. Based...
In the cause of deal with this problem optimal position and velocity tracking regarding modular reconfigurable robots (MRRs) external collisions, a zero-sum neural-optimal control method ground on critical only policy iteration (COPI) scheme is proposed by adaptive dynamic programming (ADP). The MRR dynamics model uncertainty compensated fuzzy method. Unique critic neural network (NN) (PI) ADP used as work out hamilton-Jacobi-Issacs (HJI) equation, then approximated obtained. Then Lyapunov's...
This paper presents a nonzero-sum strategy-based neuro-optimal control method for modular robot manipulators (MRMs). Based on joint torque feedback (JTF) technique, the dynamic model of manipulator systems is described as an integration subsystems. A local information-based robust compensator designed to engage uncertainty compensation, and then, optimal tracking problem MRM system transformed into n-player game issue multiple By taking advantage adaptive programming (ADP) algorithm, cost...
This paper proposes a decentralized control method for modular robot manipulators (MRMs) with external collisions based on torque estimation. Unlike some of conventional methods which depend joint sensing technique, we address the estimation and motion problems MRMs only position measurements each module. The dynamics model is established by using HD estimated torque. By local dynamic information joint, robust RBF neural network (NN) controller designed to deal uncertain collision effects...