Showing papers by "Yili Fu published in 2018"

Coordinative Motion-Based Bilateral Rehabilitation Training System with Exoskeleton and Haptic Devices for Biomedical Application.

Abstract: According to the neuro-rehabilitation theory, compared with unilateral training, bilateral training is proven to be an effective method for hemiparesis, which affects the most part of stroke patients. In this study, a novel bilateral rehabilitation training system, which incorporates a lightweight exoskeleton device worn on the affected limb; a haptic device (Phantom Premium), which is used for generating a desired tactile feedback for the affected limb; and a VR (virtual reality) graphic interface, has been developed. The use of VR technology during rehabilitation can provide goal directed tasks with rewards and motivate the patient to undertake extended rehabilitation. This paper is mainly focused on elbow joint training, and other independent joints can be trained by easily changing the VR training interface. The haptic device is adopted to enable patients to use their own decision making abilities with a tactical feedback. Integrated with a VR-based graphic interface, the goal-oriented task can help to gradually recovery their motor function with a coordinative motion between two limbs. In particular, the proposed system can accelerate neural plasticity and motor recovery in those patients with little muscle strength by using the exoskeleton device. The exoskeleton device can provide from relatively high joint impedance to near-zero impedance, and can provide a partial assist as the patient requires.

New remote centre of motion mechanism for robot-assisted minimally invasive surgery

Abstract: Robot-assisted minimally invasive surgery (RMIS) is promising for improving surgical accuracy and dexterity. As the end effector of the robotic arm, the remote centre of motion mechanism is one of the requisite terms for guaranteeing patient safety. The existing remote centre of motion mechanisms are complex and large in volume, as well as high assembly requirement and unsatisfactory precise. This paper aimed to present a new remote centre of motion mechanism for solving these problems. A new mechanism based on the RMIS requirements is proposed for holding the laparoscope and generating a remote centre of motion for the laparoscope. The mechanism kinematics is then analysed from the perspective of the structural function, and its inverse kinematics is determined with a small number of calculations. Finally, the position deviation of the laparoscope rotational point is chosen as the index to evaluate the mechanism performance. The experiments are performed to test the deviation. The position deviations of the laparoscope rotational point do not exceed 2 mm, which is lower than that of the existing remote centre of motion mechanism. The 2 mm positioning error of the laparoscope won’t affect surgeon observation of the surgical field, and the pressure caused by the positioning error was acceptable for the skin elasticity. The proposed mechanism meets the RMIS requirement. The proposed mechanism can achieve the remote centre of motion for the laparoscope. Its simple and compact structure is beneficial to avoid the collision of robotic arms, and it can be applied on other robots for providing the instrument necessary motion in minimally invasive surgery.

Variable stiffness control of series elastic actuated biped locomotion

Abstract: This study investigates the problem of dynamic walking impact on a biped robot. Two online variable stiffness control algorithms, i.e., torque balance algorithm (TBA) and surface fitting algorithm (SFA), are proposed based on virtual spring leg to achieve compliant performance. These two algorithms target on solving the high nonlinearity commonly existing in legged robot actuators. A planar biped robot experiment platform is designed for testing the proposed variable stiffness control. The experiments compare the performance of TBA and SFA and verify that applying the variable stiffness control of a virtual spring leg is capable of effectively absorbing unforeseen ground impacts and thus improving stability and safety of walking biped robots.

Joint Stiffness Identification and Deformation Compensation of Serial Robots Based on Dual Quaternion Algebra

Abstract: As the application of industrial robots is limited by low stiffness that causes low precision, a joint stiffness identification algorithm for serial robots is presented. In addition, a deformation compensation algorithm is proposed for the accuracy improvement. Both of these algorithms are formulated by dual quaternion algebra, which offers a compact, efficient, and singularity-free way in robot analysis. The joint stiffness identification algorithm is derived from stiffness modeling, which is the combination of the principle of virtual work and dual quaternion algebra. To validate the effectiveness of the proposed identification algorithm and deformation compensation algorithm, an experiment was conducted on a dual arm industrial robot SDA5F. The robot performed a drilling operation during the experiment, and the forces and torques that acted on the end-effector (EE) of both arms were measured in order to apply the deformation compensation algorithm. The results of the experiment show that the proposed identification algorithm is able to identify the joint stiffness parameters of serial industrial robots, and the deformation compensation algorithm can improve the accuracy of the position and orientation of the EE. Furthermore, the performance of the forces and torques that acted on the EE during the operation were improved as well.

Improving backdrivability in preoperative manual manipulability of minimally invasive surgery robot

Abstract: Purpose The purpose of this paper is to propose a control algorithm to improve the backdrivability performance of minimally invasive surgical robotic arms, so that precise manual manipulations of robotic arms can be performed in the preoperative operation. Design/methodology/approach First, the flexible-joint dynamic model of the 3-degree of freedom remote center motion (RCM) mechanisms of minimally invasive surgery (MIS) robot is derived and its dynamic parameters and friction parameters are identified. Next, the angular velocities and angular accelerations of joints are estimated in real time by the designed Kalman filter. Finally, a control algorithm based on Kalman filter is proposed to enhance the backdrivability of RCM mechanisms by compensating for the internally generated gravitational, frictional and inertial resistances experienced during the positioning and orientating. Findings The parameter identification for RCM mechanisms can be experimentally evaluated from comparison between the measured torques and the reconstructed torques. The accuracy and convergence of the real-time estimation of angular velocity and acceleration of the joint by the designed Kalman filter can be verified from corresponding simulation experiments. Manual adjustment experiments and animal experiments validate the effectiveness of the proposed backdrivability control algorithm. Research limitations/implications The backdrivability control algorithm presented in this paper is a universal method to enhance the manual operation performance of robots, which can be used not only in the medical robot preoperative manual manipulation but also in robot haptic interaction, industrial robot direct teaching and active rehabilitation training of rehabilitation robot and so on. Originality/value Compared with other backdrivability design methods, the proposed algorithm achieves good backdrivability for RCM mechanisms without using force sensors and accelerometers. In addition, this paper presents a new static friction compensation approach for a joint moving with very low velocity.

Automatic Extraction of the Centerline of Corpus Callosum from Segmented Mid-Sagittal MR Images.

Abstract: The centerline, as a simple and compact representation of object shape, has been used to analyze variations of the human callosal shape. However, automatic extraction of the callosal centerline remains a sophisticated problem. In this paper, we propose a method of automatic extraction of the callosal centerline from segmented mid-sagittal magnetic resonance (MR) images. A model-based point matching method is introduced to localize the anterior and posterior endpoints of the centerline. The model of the endpoint is constructed with a statistical descriptor of the shape context. Active contour modeling is adopted to drive the curve with the fixed endpoints to approximate the centerline using the gradient of the distance map of the segmented corpus callosum. Experiments with 80 segmented mid-sagittal MR images were performed. The proposed method is compared with a skeletonization method and an interactive method in terms of recovery error and reproducibility. Results indicate that the proposed method outperforms skeletonization and is comparable with and sometimes better than the interactive method.