Publications

In the research of reconnaissance robot to respond events involving hazardous materials, a novel hybrid behavior coordination mechanism based on priority and FSA is proposed. It uses a behavior group which combines several elementary behaviors based on priority to perform simple scout tasks. And then uses one of specified FSAs designed for each more complex task respectively as the behavior group selector. The key feature is that a hybrid behavior group coordinator can be structured dynamically once the corresponding task is required to be performed. Thus, such a behavior-based robot is capable of performing a goal-oriented task by this method. The implementation of a hybrid behavior coordinator used to perform the task of moving to goal is presented in detail. Simulations and experiments show the validity, robustness, and simplicity of the hybrid behavior mechanism

During robot-aided motion rehabilitation training, inappropriate difficulty of the training task usually leads the subject becoming bored or frustrated; therefore, the difficulty of the training task has an important influence on the effectiveness of training. In this study, an adaptive task level strategy is proposed to intelligently serve the subject with a task of suitable difficulty. To make the training task attractive and continuously stimulate the patient's training enthusiasm, diverse training tasks based on grabbing game with visual feedback are developed. Meanwhile, to further enhance training awareness and inculcate a sense of urgency, a dynamic score feedback method is used in the design of the training tasks. Two types of experiments, functional and clinical rehabilitation experiments, were performed with a healthy adult and two recruited stroke patients, respectively. The experimental results suggest that the proposed adaptive task level strategy and dynamic score feedback method are effective strategies with respect to incentive function and rehabilitation efficacy.

Artificial somatosensory feedback plays a crucial role in compensating for tactile and proprioceptive loss in prosthesis users. Although modern prosthetic systems can acquire rich sensory data, effectively conveying this multimodal information to the user remains a significant challenge. This study presents a wearable somatosensory feedback armband with two configurations: a multimodal version using combined vibrotactile-electrotactile (VEC) stimulation, and a unimodal version based on vibrotactile-only (VO) stimulation. In both configurations, proprioceptive feedback is conveyed via spatiotemporal vibrotactile patterns, while tactile and proximity feedback are transmitted using electrotactile stimulation in VEC and vibrotactile cues in VO. The novel system was evaluated in ten transradial amputees in psychophysical experiments, and in seven additional participants (two amputees and five non-disabled) who performed object grasping and manipulation tasks (OGMT) under four conditions. Results showed that both configurations enabled accurate recognition of multiple sensory variables, with average accuracies exceeding 90% across all conditions, and success rates above 80% in OGMT. The success rate of the proposed system was not significantly different compared to that achieved with natural visual-auditory feedback (VA). However, VA resulted in significantly lower time to perform the task. The participants reported that VEC reduced cognitive fatigue under multi-modal feedback, and VO was linked to greater willingness for long-term use. These findings demonstrate that the proposed system offers a novel, flexible, and precise platform for prosthetic sensory feedback. By leveraging multiple stimulation modalities and spatio-temporal encoding, the VEC configuration expands the range of sensory inputs, enabling more diverse, and accurate stimulation for users requiring enhanced feedback. Meanwhile, the VO configuration effectively meets most sensory feedback needs with simpler integration, making it well-suited for broader applications.

Reducing the size of magnetorheological (MR) actuators imposes severe constraints on magnetic volume and excitation efficiency, thereby limiting the achievable torque output, particularly at the 20 mm-class outer-diameter scale. To address this limitation, a miniaturized multi-drum MR actuator is proposed for enhanced torque output. It features multiple concentric annular shear gaps operating in radial shear mode and a shared magnetic flux path to improve flux utilization. Finite-element analysis is conducted to guide the magnetic circuit design, revealing a radial field distribution inherent to flux-sharing architectures and motivating an optimization strategy that drives the inner shear gap and magnetic paths close to saturation simultaneously. An optimized actuator prototype with an outer diameter of 20.8 mm, a height of 21.7 mm, and a total mass of 42.2 g was fabricated. Experimental characterization demonstrates that the actuator delivers a maximum torque of 91 mN·m with a low off-state torque of approximately 2 mN·m, achieving a substantial performance improvement over representative prior designs of comparable dimensions.

This paper describes a virtual environment system, which can produce dynamic force simulation during the control of objects in virtual environment. A 5 degrees-of-freedom haptic interface master arm with the capability to generate kinesthetic effect is combined in this system. In this system, the human operator manipulates an object in virtual environment by using the 5-DOF master arm. When virtual manipulator contacting with the virtual object, the contact force can be calculated and shown in the graphic interface. The collision response and deformation of the virtual object, which is usually called haptic rendering, also can be exhibited in graph. The experience presents an approach to improve the operator's immersion and can be used in many telerobot control fields.

Information transmission is a fundamental issue for human-computer interaction (HCI). Traditional interaction methods through visual, voice, and force haptics are very mature. However, the thermal perception (TP) for HCI is not studied in depth. This work proposes the TP-based information transmission framework. Firstly, we investigated the human hand-object contact heat transfer model and the temperature perception resolution of the hand and verified the feasibility of spatiotemporal temperature stimulation for information transmission by simulations. Then, a thermal device was designed, which utilized a 7×5 Peltier array, a water cooler, temperature sensors, and a control module to realize various static and dynamic spatiotemporal temperature patterns stimulation. Finally, we implemented a device prototype and recruited 20 subjects for experimental studies. The results show that the device can display various temperature patterns and provide thermal stimulations with high precision and speed. Furthermore, the subjects can accurately recognize different temperature values, icons, codes, and waveforms with their palm and fingers after a few times of training, which validates the TP-based information transmission method. Therefore, people can apply this method to interact with machines for information feedback, virtual reality, augmented reality, etc.

非接触式传感器相比接触式触觉传感器可以避免与物体直接接触过程中产生的噪声, 因而能够获取更有价值的原始数据表征物体内在属性; 然而针对非接触式传感器感知的物体属性数据而言, 现有算法难以实现单样本学习下的物体准确识别. 为解决这一问题, 本文提出一种新颖的单样本学习下时序约束稀疏表示方法（One-Shot Learning with Temporally Constrained Sparse Representation, OSL-TCSR）用于识别5种不同材料下的50个物体类别. 首先将两种原始数据（Lumini光谱和SCiO光谱）并行投影至共享子空间, 并且使用聚类典型关联分析法（C-CCA）计算两种原始数据的聚类相关性特征; 其次通过字典学习分别计算得到的聚类相关性特征数据以及原始数据的编码向量, 并利用原始数据的编码向量对相关性特征数据的编码向量进行二次投影映射; 然后通过将两次映射后的原始数据和相关性特征数据进行重构, 以充分耦合化两种光谱数据; 进一步地, 通过设计新颖的时序约束正则化的稀疏表示方法计算重构后的原始数据和相关性特征数据, 以充分考虑每个光谱序列的时序特征; 最后通过与最新的物体识别方法进行实验对比, 结果表明提出的OSL-TCSR方法提高了单样本学习情况下的物体识别结果. 此外, OSL-TCSR还可灵活迁移至多种应用场景, 比如材料识别或纹理识别等.

The communication time delay and uncertain models of robotic manipulators are the major problem in the teleoperation system, which can reduce the performance and stability of the system. This article proposed a novel finite-time adaptive control scheme for position and force tracking performances of the teleoperation system. First, a combined auxiliary error system with position and force tracking errors is designed. Second, a velocity feedback filter is introduced, and a new auxiliary variable function with finite-time structure is designed for controller design. The radial basis function neural network (RBFNN) is applied to estimate the uncertain parts. Then the finite-time adaptive control scheme and adaptive laws are given. Third, based on the Lyapunov method, stability and finite-time performance are demonstrated. And finally, the simulation and experimental studies (with Phantom Ommi devices) are performed and demonstrate the effectiveness of the proposed control scheme on teleoperation position/force tracking.

It is a challenging task for operators to interact with the remote environment without its geometric and dynamic knowledge during teleoperation. In this article, a novel system architecture for implementing the translational object modeling and correction during remote interaction is proposed to reconstruct the haptic interaction and predict the object motion at the local virtual reality–based teleoperation. First, a stress mutation analysis method is proposed for segmenting the translational object motion into static phase, critical phase, and sliding phase. And the static limiting friction is originally estimated in the teleoperation area. Meanwhile, mass-damper-spring model and adapted Karnopp friction model are adopted for dynamic modeling in each phase. Second, a novel adaptive forgetting factor recursive least square method is studied for high-accuracy parameter estimation. With the estimated model parameters, the motion of the translational object is predicted at the master side. Meanwhile, for model consistence between the real and virtual environments, a new correction strategy is used to adaptively update the environment model. According to the experimental results, the translational object can be accurately modeled in real time, and its motion at the master side can be predicted precisely and corrected promptly.