Article
Engineering, Multidisciplinary
Wafa Znegui, Hassene Gritli, Safya Belghith
Summary: This paper aims to develop an improved closed-form analytical expression for analyzing the complex walking behavior and stability of a passive-dynamics biped walker. By linearizing and approximating the hybrid dynamics, an enhanced closed-form expression of the Poincaré map and an analytical expression for the computation of step period are designed. A simplification of these expressions is achieved by decreasing dimension and providing a Jacobian matrix expression for investigating the stability of the designed simplified Poincaré map.
APPLIED MATHEMATICAL MODELLING
(2021)
Article
Mathematics, Applied
Lizhi Niu, Wei Xu, Qin Guo
Summary: This paper investigates the transient response of time-delay systems under Gaussian white noise and explores the effectiveness of the complex fractional moments method. The study uses approximate transformation and stochastic averaging methods to derive one-dimensional stochastic differential equations and complex fractional moments, providing insights into system behavior and noise intensity effects. Furthermore, the application of the complex fractional moments method to analyze transient bifurcation and discuss the imperfect symmetry property adds new perspectives to the research field.
Article
Multidisciplinary Sciences
Guanfeng Zhou, Bo Jiang, Tengfei Long, Guirong Jiang
Summary: This article presents a biped robot walking on horizontal ground with two feasible switching patterns of motion (two-phase gait and three-phase gait). By using the first-order Taylor approximate at the equilibrium point, a simplified linear continuous dynamic equation is obtained to discuss the walking dynamics of the biped robot. Conditions for the existence and stability of period-1 gaits (P(1,2),P(1,3)) and period-2 gaits (P(2,2,2),P(2,2,3),P(2,3,3)) are obtained by using a discrete map. Among the periodic gaits, the P(2,2,3) type gait has never been reported in previous studies. Flip bifurcation of periodic gait is investigated. Numerical results for periodic gaits and bifurcation diagram are in good agreement with the theoretical analysis.
PROCEEDINGS OF THE ROYAL SOCIETY A-MATHEMATICAL PHYSICAL AND ENGINEERING SCIENCES
(2023)
Article
Engineering, Mechanical
Masoumeh Safartoobi, Morteza Dardel, Hamidreza Mohammadi Daniali
Summary: This paper presents the application of an internal damping mechanism to the mathematical description of the simplest passive walking model, making walking natural and efficient. By using viscoelastic legs and precise boundary conditions, the biped starts walking from a stable condition and exhibits stable period-one gait cycles. Numerical simulations demonstrate that the overall effect of viscoelastic legs on passive walking is efficient in terms of stability and energy dissipation.
NONLINEAR DYNAMICS
(2022)
Article
Biology
Nicole Y. Lai, Jordan M. Bell, David Bodznick
Summary: The process of movement in animals generates sensory stimulation known as reafference, which needs to be selectively filtered in order to detect relevant external sensory signals. In elasmobranch fish, an adaptive filter in the cerebellum-like electrosensory nucleus generates cancellation signals to suppress predictable reafference, utilizing behavior-associated internal reference signals. Multiple streams of internal reference signals result in specific cancellation signals that are predictive of sensory reafference. Individual AENs can form multiple cancellation signals simultaneously and update them incrementally during continuous behaviors, demonstrating the adaptive filter's capacity to suppress dynamic patterns of reafference arising from complex co-occurring behaviors.
JOURNAL OF EXPERIMENTAL BIOLOGY
(2021)
Article
Automation & Control Systems
Hae Yeon Park, Jung Hoon Kim, Ko Yamamoto
Summary: This article proposes a new stability criterion for biped walking systems on the linear inverted pendulum model, and introduces a norm-based criterion and related control approaches. The research shows that the stability regions can be explicitly obtained through only finite numbers of computations.
IEEE TRANSACTIONS ON INDUSTRIAL INFORMATICS
(2022)
Article
Engineering, Electrical & Electronic
Aiwen Luo, Sandip Bhattacharya, Sunandan Dutta, Yoshihiro Ochi, Mitiko Miura-Mattausch, Jian Weng, Yicong Zhou, Hans J. Mattausch
Summary: The study aims to implement surface recognition through walking-pattern classification using a support vector machine (SVM) and time-domain feature descriptors. By extracting real-time dynamic force-sensing data streams and applying multiple binary SVM classifiers, the developed cost-efficient and accurate surface-recognition system is valuable for ensuring safe indoor locomotion for biped robots and enhancing their understanding of the human living environment.
IEEE SENSORS JOURNAL
(2021)
Article
Engineering, Multidisciplinary
Nicholas S. Szczecinski, Chris J. Dallmann, Roger D. Quinn, Sasha N. Zill
Summary: Control of forces is crucial for animals and walking machines. Insects measure forces through campaniform sensilla (CS) in their exoskeletons, with CS afferent firing frequency being highly dynamic and adapting to different forces over time. A new phenomenological model predicts CS discharge based on the instantaneous stimulus force relative to an adaptive variable, which can replicate responses to various force stimuli and exhibit power-law adaptation.
BIOINSPIRATION & BIOMIMETICS
(2021)
Article
Engineering, Mechanical
Masoumeh Safartoobi, Morteza Dardel, Hamidreza Mohammadi Daniali
Summary: This paper investigates a passive walking robot model with flexible legs, utilizes numerical techniques to find suitable initial conditions for gait cycles, and demonstrates that elastic legs result in non-periodic motion patterns for small slope angles.
MECHANISM AND MACHINE THEORY
(2021)
Article
Chemistry, Analytical
Xuanyang Shi, Junyao Gao, Yizhou Lu, Dingkui Tian, Yi Liu
Summary: The paper introduces a method of real-time optimization using a CoM-velocity feedback controller, constructing a touchdown return map and using nonlinear least squares to optimize spring stiffness. Hierarchical quadratic programming optimization is used for strict task priority, with the dynamic equation given the highest priority, reducing the number of optimizations.
Article
Computer Science, Information Systems
Gyoo-Chul Kang, Yoonsang Lee
Summary: This paper presents a deep reinforcement learning method based on finite state machines for character control without the need for motion data. The control policy learns to output target poses and transition timings between states, allowing the simulated character to walk according to desired gait parameters.
Article
Robotics
Yisoo Lee, Hosang Lee, Jinoh Lee, Jaeheung Park
Summary: The study develops an event-based finite-state machine (E-FSM) to enhance reactivity of biped robots to unforeseen disturbances for robust walking. The proposed control method ensures stable locomotion under external disturbances, as validated through experiments.
IEEE TRANSACTIONS ON ROBOTICS
(2022)
Article
Engineering, Biomedical
Lauren Chee, Giacomo Valle, Michele Marazzi, Greta Preatoni, Florian L. Haufe, Michele Xiloyannis, Robert Riener, Stanisa Raspopovic
Summary: This study presents a non-invasive neurostimulation system that can encode foot-ground and knee angle information, leading to decreased metabolic consumption and increased prosthesis confidence in lower-limb amputees. The system also promotes more physiological walking strategies and increased stability in response to external perturbations.
JOURNAL OF NEURAL ENGINEERING
(2022)
Article
Mathematics, Applied
Minjuan Yuan, Liang Wang, Yiyu Jiao, Wei Xu
Summary: In this paper, an improved cell mapping method based on dimension-extension is proposed to study the global dynamics of fractional systems, which successfully describes one-step mappings using Markov chains, avoiding memory losses. Simulations demonstrate the high accuracy and efficiency of the method, revealing abundant global dynamics behaviors in fractional smooth and discontinuous oscillators.
Article
Engineering, Biomedical
Romain Valette, Jose Gonzalez-Vargas, Strahinja Dosen
Summary: This study evaluated the psychometric properties of multichannel electrotactile stimulation during sitting and walking. The results showed that sitting provided better perception of the stimulation compared to walking, which can be affected by motion and physical coupling. Calibration of the stimulation is particularly important for individuals with lower-limb loss.
JOURNAL OF NEUROENGINEERING AND REHABILITATION
(2023)
Article
Automation & Control Systems
Lianxin Yang, Caihua Xiong, Ming Hao, Yuquan Leng, Ken Chen, Chenglong Fu
Summary: The study aims to investigate human response to different patterns of relative load movement, providing guidance for controlling the actively suspended backpack.
IEEE-ASME TRANSACTIONS ON MECHATRONICS
(2022)
Article
Engineering, Multidisciplinary
Yixuan Guo, Zhouxiang Jiang, Bao Song, Xiaoqi Tang, Huasong Min, Chenglong Fu
Summary: This paper proposes a kinematic calibration method based on an optimized distance error function to improve the accuracy of relative positions in industrial robots. By designing a three-dimensional measurement system and an optimization function, the robot calibration is successfully implemented, and the influence of redundant parameters on relative position errors is analyzed.
Article
Robotics
Xinxing Chen, Bo Yang, Jian Huang, Yuquan Leng, Chenglong Fu
Summary: In this paper, a multi-continuous-output fuzzy system was designed and tuned with reinforcement learning to solve the robotic odor source localization problem in dynamic odor plumes. The proposed method showed good performance in terms of efficiency and success rate, as validated through simulations and real robotic experiments.
Article
Automation & Control Systems
Yuepeng Qian, Shuaishuai Han, Yining Wang, Haoyong Yu, Chenglong Fu
Summary: This article introduces a novel nonlinear series elastic actuator (nSEA) for transparent actuation and safe human-robot interaction in assistive exoskeletons. The nSEA features a optimized nonlinear series elastic element and a quasi-direct drive motor, which results in low mechanical impedance, high backdrivability, and less acoustic noise. A new torque control method based on cascade PI control is proposed for accurate and robust interaction torque control. Experimental evaluation with human subjects validates the advantages of the nSEA-driven hip exoskeleton in achieving actuation transparency and safety.
IEEE-ASME TRANSACTIONS ON MECHATRONICS
(2023)
Article
Computer Science, Artificial Intelligence
Kuangen Zhang, Jiahong Chen, Jing Wang, Xinxing Chen, Yuquan Leng, Clarence W. de Silva, Chenglong Fu
Summary: This paper proposes an ensemble diverse hypotheses and knowledge distillation method to address the issues in human intent prediction and human activity recognition. The method reduces cross-subject divergence and learns diverse and complete features by training feature generators. A lightweight student network distills the knowledge from the ensemble model to decrease the model size while maintaining accuracy.
INFORMATION FUSION
(2023)
Article
Engineering, Electrical & Electronic
Ying Zang, Chenglong Fu, Dongsheng Yang, Hui Li, Chaotao Ding, Qingshan Liu
Summary: Multispectral pedestrian detection has attracted great attention due to the improved performance it offers. However, existing algorithms still face issues such as inadequate information exchange and lack of targeted network design. To address this, we propose two subnetworks, FTHd and FTn, for daytime and nighttime images, respectively. We emphasize the texture features of RGB images during the day by adding a histogram layer, while during the night, we combine the two streams to reduce computation and leverage the CFT module for feature fusion and interaction. Our FTHd and FTn achieve improved pedestrian detection accuracy compared to baseline methods.
SIGNAL IMAGE AND VIDEO PROCESSING
(2023)
Article
Robotics
Chengjie Zhang, Chengyu Lin, Yuquan Leng, Zezheng Fu, Yaoyu Cheng, Chenglong Fu
Summary: This study proposes an arm-free HRI based on Mixed Reality (MR) feedback and head control, which enables effective control of the robotic arm for grasping densely placed objects. The MR feedback provides 3D visualization of robot perception results and aligns virtual grippers with real grippers. Head control allows disabled individuals to flexibly and effectively control the gripper's 6D pose.
IEEE ROBOTICS AND AUTOMATION LETTERS
(2023)
Article
Engineering, Environmental
Chenglong Fu, Changmei Lin, Weipeng Zhang, Yaling Lin, Jinhong Xiu, Yonghao Ni, Liulian Huang
Summary: This study reports a simple process for extracting microfibrillated cellulose fibers from conventional bleached kraft pulp, which is particularly suitable for flexible electronics applications. The obtained fibers have a high specific surface area, resulting in favorable properties such as high transparency and tensile strength.
CHEMICAL ENGINEERING JOURNAL
(2023)
Article
Automation & Control Systems
Jiyong Tan, Bing Li, Yuanwei Li, Bin Li, Xinxing Chen, Jiayi Wu, Baoming Luo, Yuquan Leng, Yiming Rong, Chenglong Fu
Summary: The study proposes a flexible and fully autonomous ultrasound scanning system for breast ultrasound imaging, which effectively improves the quality and stability of ultrasound images by controlling the contact force. Experimental verification shows the feasibility and superiority of the system.
IEEE TRANSACTIONS ON AUTOMATION SCIENCE AND ENGINEERING
(2023)
Article
Robotics
Chengyu Lin, Chengjie Zhang, Xiaoyu Yan, Yuquan Leng, Chenglong Fu
Summary: This letter proposes a Voluntary-Redundant hybrid control (VRHC) method to achieve multiple degrees of freedom (DoFs) control of SuperLimb by integrating the user's voluntary movement and redundant muscle signals. Nine intuitive redundant muscle patterns (IRMPs) are identified by a convolutional neural network, and a muscle-velocity mapping method based on redundant muscle force is proposed to control the velocity of the SuperLimb. The results show that the proposed method provides an intuitive velocity control with multiple DoFs for SuperLimb to assist the user in complex coordination tasks.
IEEE ROBOTICS AND AUTOMATION LETTERS
(2023)
Proceedings Paper
Computer Science, Artificial Intelligence
Tianrun Chen, Chenglong Fu, Ying Zang, Lanyun Zhu, Jia Zhang, Papa Mao, Lingyun Sun
Summary: The rapid development of AR/VR brings high demands for 3D content. Sketch-based 3D modeling offers a potential solution but faces challenges with sparse and ambiguous sketches. This work introduces an end-to-end approach, Deep3DSketch+, that performs 3D modeling using a single free-hand sketch, achieving high-fidelity performance through a lightweight generation network and structural-aware adversarial training with a Stroke Enhancement Module (SEM).
MULTIMEDIA MODELING, MMM 2023, PT II
(2023)
Article
Automation & Control Systems
Jiyong Tan, Jiawang Li, Yuanwei Li, Bing Li, Yuquan Leng, Yiming Rong, Chenglong Fu
Summary: Ultrasound image guidance provides real-time display of soft tissue movement, which is crucial for tumor examinations and interventions. This paper proposes a real-time autonomous ultrasound robot trajectory planning framework for global scanning of breast tissue and real-time local tracking of suspicious tumor targets.
IEEE TRANSACTIONS ON AUTOMATION SCIENCE AND ENGINEERING
(2023)
Article
Engineering, Biomedical
Yuepeng Qian, Haoyong Yu, Chenglong Fu
Summary: Gait asymmetry is a common clinical characteristic of stroke patients that negatively impacts their quality of life. Robot assistance has been shown to improve functional walking in stroke patients, but few studies have focused on correcting gait asymmetry. This study developed an adaptive oscillator-based assistive torque control system that corrected gait asymmetry at the hip joints during the entire gait cycle.
IEEE TRANSACTIONS ON NEURAL SYSTEMS AND REHABILITATION ENGINEERING
(2022)
Proceedings Paper
Computer Science, Hardware & Architecture
Chenglong Fu, Qiang Zeng, Haotian Chi, Xiaojiang Du, Siva Likitha Valluru
Summary: This paper reveals a new type of attack against Internet of Things automation systems, using event message delay and command message delay as attack primitives. By compromising one Wi-Fi device, an attacker can delay the messages of other IoT devices, which is called IoT Phantom-Delay Attacks. This study demonstrates that this attack can be used to launch various attacks on vulnerable IoT devices.
2022 52ND ANNUAL IEEE/IFIP INTERNATIONAL CONFERENCE ON DEPENDABLE SYSTEMS AND NETWORKS (DSN 2022)
(2022)
Article
Engineering, Biomedical
Chuheng Chen, Kuangen Zhang, Yuquan Leng, Xinxing Chen, Chenglong Fu
Summary: This paper presents an unsupervised sim-to-real adaptation method to accurately classify common real-world environments and assist amputee's terrain-adaptive locomotion. By training the network on simulated data and unlabeled real-world data, and visualizing the essential features, the proposed method achieves accurate environmental classification.
IEEE TRANSACTIONS ON NEURAL SYSTEMS AND REHABILITATION ENGINEERING
(2022)