736 publications from this institution
The proposed ESD method combining SE with SVs and SS can serve as an assistive diagnostic tool to help medical professionals and individuals detect and prevent strokes at an earlier stage, reduce workload, and improve identification objectivity. The code and experimental protocol of this paper are available at https://github.com/LiuYingchenseu/ESPVC.
This paper presents a novel parallel rhombus-chain-connected haptic deformation model based on physics. Softness haptic deformation model with high precision and real-time calculation property is a key issue for interaction between human and virtual reality. In our model, when a certain chain structure unit is pushed or pulled by an external force, chain connected structure between every unit will make the chain structure units which are connected one by one move and produce deformation synchronously. Despite the haptic refresh rate requirement of more than 200 Hz, our parallel rhombus-chain-connected model satisfies real-time performance because the simple nature of the model, so that it can be convenient and fast for computation. Experiments are conducted with virtual human liver model, the result demonstrates that our model provides not only stable, but also realistic haptic feeling to user.
In this paper, we describe a heading direction correction algorithm for a tracked mobile robot. To save hardware resources as far as possible, the mobile robot's wrist camera is used as the only sensor, which is rotated to face stairs. An ensemble heading deviation detector is proposed to help the mobile robot correct its heading direction. To improve the generalization ability, a multi-scale Gabor filter is used to process the input image previously. Final deviation result is acquired by applying the majority vote strategy on all the classifiers' results. The experimental results show that our detector is able to enable the mobile robot to correct its heading direction adaptively while it is climbing the stairs.
Telerobot system plays an important role in executing task under hazard environment. As the computer networks such as the Internet are being used as the communication channel of telerobot systems, varying time delay causes the overall system unstable and reduces the performance of transparency. In this paper, we propose twelve operation modes with different control schemes for telerobot on the Internet with time delay. And an optimal operation mode with control scheme is specified for telerobot with time delay, based on the tradeoff between passivity and transparency properties. We experimentally confirm the validity of the proposed optimal mode and control scheme by using a simple one DOF master–slave manipulator system.
The constrained command tracking problem for the cloud robotic system with unknown bounded time-varying delays is considered. Based on a novel predictor-observer approach, a command governor is proposed to modify the received command in order to guarantee the constraint enforcement of remote robot. We firstly use a prediction algorithm in the forward channel to generate a virtual command, which is close to the original one as much as possible when the danger of constraint violation exists. Then, to deal with the measurement delay and disturbance in the backward channel, interval estimations of the system states are applied instead of delayed measurements in the prediction algorithm of the command governor. The constraints can be rigorously enforced with the upper and lower estimations. The efficiency of the proposed approach is demonstrated by simulations in the case of a single-degree-of-freedom manipulator.
A genetic algorithm and recursive least squares (RLS) learning algorithm for a Gaussian radial basis function network is described, for modelling and predicting nonlinear time series. Better generalisation performance can be achieved than that of the usual clustering and RLS method.
Two aspects of enhancing the robot’s autonomous capabilities and improving the manner of human-robot interaction are addressed in the research and development of reconnaissance robot to respond events involving hazardous materials. A novel human-robot collaborative semi-autonomous mobile robot architecture (SAMRA) which combined the key advances of deliberation and reactivity architecture is proposed. In which, the human operator collaborate with the robot to perform tasks. The implement method of primary intelligent behaviors which use fuzzy inference theory and dynamic priority are presented in succession. And then, a visualized human-robot interface with lively video display, 3D drawing simulator of robot’s attitude and manipulator’s pose is introduced briefly. Simulation and experimental tests show the human-robot collaborative teleoperation system of reconnaissance robot possesses of flexibility and robustness in the performing tasks.
Brain–machine fusion, also known as hybrid intelligence or brain–computer interface (BCI), is considered one of the most promising technologies of the 21st century. Its potential impact spans a wide range of disciplines, including cognitive science, information science, artificial intelligence, biology, neuroscience, and engineering. The research in this field aims to seamlessly integrate biological intelligence (i.e., the human brain) with machine intelligence (computers or robots) to create a new, powerful form of hybrid intelligence that far surpasses the limitations of current biological and machine intelligence systems. Brain–machine fusion not only signifies the convergence of cutting-edge science and technology but also heralds a new era in which the way humans interact with machines undergoes a profound transformation. The research in this field delves deep into the understanding of human thought processes and cognition, as well as the creation of novel sensory and motor channels to facilitate more natural and intuitive interactions. The scope of brain–machine fusion research extends beyond mere information exchange, encompassing the integration of emotions and motivations. Understanding and interpreting a user’s emotional state and motivations are crucial for optimizing the performance of fusion systems, aiding in better meeting user needs and providing more personalized experiences. A key objective is enhancing a user’s operational capacity in handling complex tasks. This can encompass highly intricate decision making, problem solving, and task execution, with broad applications in fields, such as healthcare, military, industry, and entertainment. Furthermore, brain–machine fusion necessitates the development of cognitive interaction models that can adapt actively to a user’s cognitive characteristics and integrate with machine learning algorithms to achieve personalized adaptability in intelligent systems, thereby enhancing the level of interaction between the user and the system.
This paper presents a spherical actuator-based hand-held device that provides lateral force feedback for user interaction with the touch screen. The spherical actuator is built around a ball containing the magnetorheological elastomer (MRE). This design of the actuator not only allows it to move in multiple degrees of freedom but also realizes its direct interaction with the touch screen and enhances its lateral force feedback capability by utilizing the conductive and soft characteristics of the MRE. Meanwhile, using the magnetically conductive properties of the MRE, the lateral force can be controlled by current. In this paper, we introduced the overall structure of the device, described the fabrication of the MRE, and tested the relative permeability and surface friction properties of the MRE. Then, based on the structural parameters obtained by lateral force modeling and finite element analysis, we fabricated a prototype of the actuator and determined the lateral force control method through calibration tests. Finally, through physical measurements and psychophysical experiment, we comprehensively evaluated the lateral force performance of the actuator and its ability in displaying virtual surface friction. The experimental results confirm the effectiveness of the actuator in interacting with the touch screen and displaying the virtual surface friction characteristics.
