Publications

The weightless environment causes difficulties for anchoring and sampling in asteroid exploration. In this paper, we present the analyses of the anchoring and sampling processes of a landing robot to deal with the difficulties. The mechanism and electrical system of the robot are introduced. The robot is composed of a body, three landing legs, three anchoring legs, one sampling arm, three reverse thrusters, a sensing system, and a control system. Diamond cutting discs are installed at the end of the anchoring legs and sampling arm for penetrating the rock surface of asteroids in anchoring and sampling. After anchoring on the surface by using the anchoring legs, the robot can sample hard rock material via the sampling arm. The robot could sample a 13.86 cm <sup xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">3</sup> rock piece at one time if the diameter of the cutting disc is 6 cm. The motion control results of the anchoring leg are also given. The results of this paper show the possibility of anchoring and sampling by using only one robot.

This paper presents a new design and realization of an array tactile sensor for artery pulse measurement. In this array tactile sensor, five pieces of PVDF films are distributed in a circle as sensing elements. Among them one PVDF is located in centre and others are located surrounding the centre. This circular array structure ensure the sensor can not only measure signals of the strong pulse point in the centre, but also show the transfer effect of pulse signals from the centre point to the surrounding points in all directions, which is convenient for artery pulse wave analysis. The five-path signal conditioning circuits and the USB interface circuit are designed. This sensor can measure human pulse from multiple points and display the waveforms of the center measurement point, together with its surrounding circumstances by interactive software. It is very suitable for the next generation of pulse waveform analysis and lays the foundation for medical diagnosis.

Vision-guided telerobot is often used to execute tasks, such as grasping and classification, in various environments, which contains some unfamiliar objects beyond its matching library. Hence, it is necessary to create new template dynamically for the unfamiliar objects. However, this procedure is inconvenient for the traditional template matching algorithm. In this article, a novel map–based normalized cross correlation algorithm is proposed. Map–based normalized cross correlation is summarized into two phases. In the learning phase, map–based normalized cross correlation creates new template and map by the superpixel-based GrabCut method dynamically, which is different from previous template matching algorithms. In the matching phase, a map-based similarity evaluation is designed to determine the position and rotation angle of object, where the map is used to eliminate the interference of background. Various experiments demonstrate that superpixel-based GrabCut method is more robust against noise than the traditional GrabCut algorithm and can separate the object from texture-rich background with less iteration times and time consumption. Additionally, map–based normalized cross correlation algorithm can locate objects in texture-rich images more accurately compared with polar transformation and image pyramids normalized cross correlation algorithm, especially for the matching of irregularly shaped object.

It is a difficult task to teleoperate a robot in a partially known or unstructured environment without any assistance.In this paper, a haptic assisted teleoperation system is discussed that provides virtual fixtures for motion guidance and haptic feedbacks for dynamic interaction.First, a novel virtual fixture generation method based on a point-set implicit surface was proposed in a point cloud augmented virtual environment for human operational guidance.A robot-centered potential force field model was applied to generate guidance virtual fixtures.The resultant forces generated from both forbidden region and guidance virtual fixtures were fed back in real time to the human operator through a haptic device.Second, a real-time dynamic modelling method was proposed to reconstruct contacts in the teleoperation system.For the reconstruction of dynamic properties, an adaptive forgetting factor recursive least-squares method was applied for real-time parameter estimation.With haptic assistance from the motion guidance force and the dynamic force feedback, human operators could efficiently achieve targets and complete interactive tasks.The experimental results show that the proposed methods are effective for robot teleoperation.Sensors and Materials, Vol. 29, No. 9 (2017) implemented VF using definite surfaces to assist in peg-in-hole tasks and thereby increased operator performance up to 70%.Haptic VFs are mainly categorized into two types, namely, forbidden region virtual fixtures (FRVFs) and guidance virtual fixtures (GVFs).These are systemgenerated forces that are fed back to the operators as motion regulation during robot manipulation.As the names imply, the FRVFs are used to restrict robot access to "forbidden" regions, while the GVFs assist operators or robots to move along desired paths or towards targets.The VFs are widely applied in robotic surgery, autonomous robotic manipulators, and robot teleoperation. (5)he potential benefits of VFs are that it is safer and faster to operate robots with them.In addition, VFs can reduce mental workload, time on task, and errors.VFs have been widely studied.Abbott focused mainly on control stability in classical teleoperation control architectures with VFs. (6)Park and Howard (7) proposed a methodology that employs vision-based GVF techniques for improving human performance in a teleoperated manipulation system.Kapoor and Taylor (8) introduced the notion of "soft" virtual fixture mechanisms for robotic surgical assistance.Most early reported studies concentrated on system performance using VFs (9,10) or VF applications based on a predefined geometric surface or path. (11,12)ore VF construction methods have been described in a recent survey. (5)Although these studies successfully implemented VFs, using VFs in a partially known and unstructured environment is still challenging.Recently, VF construction based on computer vision has been widely studied for adaptive applications, particularly for use in dynamic and unstructured environments.Yamamoto et al. (13) applied FRVFs to tissues based on shape recognition.However, this approach was offline and not adaptive.A construction method for real-time FRVFs during teleoperation from streaming point clouds obtained using an RGB-D camera has been proposed by Kosari et al. for robot teleoperation. (14)owever, they focused primarily on FRVFs for three architectures used in teleoperation.Furthermore, when human operators control the robot into contact with a real target, the dynamic properties of the environment are often hard to feed back to the VE.Estimating the dynamic properties of the contact between the robot and the object is essential for haptic rendering in the VE.Therefore, if the dynamic model can perfectly describe the real environment, human operators can perceive the virtual contact force directly in the VE during manipulation.However, implementing accurate dynamic modelling remains a challenge.Many studies have been done on the handling of these problems in robotic applications.Ni et al. (15) proposed a sliding-average least-squares algorithm-based environment identification method for contact interaction with

The estimation of the grip force and the 3D push-pull force (push and pull force in the three dimension space) from the electromyogram (EMG) signal is of great importance in the dexterous control of the EMG prosthetic hand. In this paper, an action force estimation method which is based on the eight channels of the surface EMG (sEMG) and the Generalized Regression Neural Network (GRNN) is proposed to meet the requirements of the force control of the intelligent EMG prosthetic hand. Firstly, the experimental platform, the acquisition of the sEMG, the feature extraction of the sEMG and the construction of GRNN are described. Then, the multi-channels of the sEMG when the hand is moving are captured by the EMG sensors attached on eight different positions of the arm skin surface. Meanwhile, a grip force sensor and a three dimension force sensor are adopted to measure the output force of the human's hand. The characteristic matrix of the sEMG and the force signals are used to construct the GRNN. The mean absolute value and the root mean square of the estimation errors, the correlation coefficients between the actual force and the estimated force are employed to assess the accuracy of the estimation. Analysis of variance (ANOVA) is also employed to test the difference of the force estimation. The experiments are implemented to verify the effectiveness of the proposed estimation method and the results show that the output force of the human's hand can be correctly estimated by using sEMG and GRNN method.