Intelligent Robotics and Applications

To efficiently facilitate autonomy sharing for assisting mobile robot teleoperation, in this paper we propose a method to recognize four contextual task types executed by the human operator: doorway crossing, object inspection, wall following and robot docking, which extends our previous approach, where only the first two task types were considered. We employ a set of simple but highly distinctive task features to efficiently describe each task type, which is adopted by a Gaussian Mixture Regression (GMR) model combined with a recursive Bayesian filter (RBF) to infer the most probable task the human operator executes across multiple candidates during operation. We demonstrate the effectiveness of the approach with a variety of tests in a cluttered indoor scenario.

This paper introduced a novel approach to recognize objects with tactile images by utilizing semi-supervised learning approaches. In tactile object recognition, the data are normally insufficient to build robust training models. Thus the model of Ensemble Manifold Regularization, which combines concepts of multi-view learning and semi-supervised learning, is adapted in tactile sensing to achieve better recognition accuracy. Different outputs of classic bag of words with different dictionary sizes are considered as different views to produce an optimized one based on multiple graphs learning optimization. In the experiments 12 objects were used to compare the classification performances of our proposed approach and the classic BoW model and it is proved that our proposed method outperforms the classic BoW framework and objects with similar features can be better classified.

Along with the progress of robot technology, nowadays industrial robot manipulators are applied for more diverse and complex tasks. Consequently, the conventional manipulative devices, such as the teaching pendant and keyboard, become unsuitable for task teaching or planning. Motivated by it, in this paper, we propose a new type of manipulation system for industrial robot manipulators based on tablet PC. With wireless communication, the proposed system provides flexibility for remote manipulation. In addition, a graphical interface and several assistive tools are equipped for viewing and maneuvering tasks during execution, allowing the operator to conduct the task in a more intuitive way. We conducted experiments using both the proposed manipulation system and conventional manipulative device.

The research presented in this paper is part of an ongoing research project which looks at the role cognitive bias plays in developing long-term companionship between a cognitively imperfect robot and humans. In this paper we discuss, how a ‘human-like’ trait such as

forgetfulness

can play a role in robot-human interaction to develop long-term companionship. One of the robots used in this study called ERWIN was given a level of

forgetfulness

as a reference of

Misattribution

for its cognitive bias. It is hoped that humanlike fallible characteristics (e.g. making mistakes, wrong selection, forgetfulness and other imperfect behaviours) can help in developing a more natural and believable attachment bond between Robots and Humans. By developing forgetfulness in a robot, it is expected that the user can relate to the robot more easily which in turn can help them to develop a stronger long-term companionship towards the robot. In previous experiments the robot ERWIN

forgot

and

misattributed

some of the participant’s information which was noted previously. The experimental results show that the participants initially warmed to the robot with the forgetfulness trait. To continue experimenting with misattribution and other cognitive biases, current experiments with a humanoid robot MARC and a group of participants is being carried out.

Close natural interaction is assumed to be a key feature for a socially assistive service robot, if the user is expected to develop an emotional bond to the robot system and accept it as a companion over a prolonged period. One particularly intuitive way of affective interaction, seen e.g. between people and pets, is physical touch in the form of stroking or smacking. In order to support the companion robot’s role of an intelligent pet-like buddy, we aim to equip the robot with the capabilities to recognize such physical interaction behavior and react accordingly. In this paper, we present a low cost smart fur sensor which encourages tactile interaction by the user and recognizes different touch patterns relating to various kinds of emotional expressions. Special emphasis is put on the simple and robust classification of touch patterns emerging from such tactile interaction to distinguish the respective inputs for the robot.

We present a robotic tool that autonomously follows a conversation to enable remote presence in video conferencing. When humans participate in a meeting with the help of video conferencing tools, it is crucial that they are able to follow the conversation both with acoustic and visual input. To this end, we design and implement a video conferencing tool robot that uses binaural sound source localization as its main source to autonomously orient towards the currently talking speaker. To increase robustness of the acoustic cue against noise we supplement the sound localization with a source detection stage. Also, we include a simple onset detector to retain fast response times. Since we only use two microphones, we are confronted with ambiguities on whether a source is in front or behind the device. We resolve these ambiguities with the help of face detection and additional moves. We tailor the system to our target scenarios in experiments with a four minute scripted conversation. In these experiments we evaluate the influence of different system settings on the responsiveness and accuracy of the device.

This work describes the development of flexible tactile sensor shoe inlays for humanoid robots. Their design is based on a sandwich structure of flexible layers with a thin sheet of piezoresistive rubber as main transducer element. The layout and patterning of top and bottom electrodes give 1024 pressure sensitive cells and the use of high speed electronics and multiplexing algorithms provides frame rates of 100 Hz. The sensors tolerate overloads while showing a consistent output. The developed prototypes show a high potential not only for robotics, but also for use in sensorised human prosthetics.

The existing procedures for Autism Spectrum Disorder (ASD) diagnosis are often time consuming and tiresome both for highly-trained human evaluators and children. In addition, prospective human evaluators need to undergo a rigorous and lengthy training process that may not be accessible or affordable to all interested individuals. Hence, this paper proposes a framework for robot-assisted ASD evaluation based on Partially Observable Markov Decision Process (POMDP) modelling. POMDP is broadly used for modelling optimal sequential decision making tasks under uncertainty. Spurred by the widely accepted Autism Diagnostic Observation Schedule (ADOS), we start off with emulating ADOS. In other words, our POMDP model explicitly takes into account the ADOS stratification into several modules, ongoing task informativeness and robotic sensor deficiencies. Relying only on imperfect sensor observations, the robot provides an assessment of the child’s ASD-relevant functioning level (which is partially observable) within a particular task. Finally, we demonstrate that the proposed POMDP framework provides fine-grained outcome quantification, which could also increase the appeal of robot-assisted diagnostic protocols in the future.

Human-computer interface (HCI) is an important way for information transmission between human and computer. This study aims to design a kind of noncontact HCI. Handwriting recognition for computer input is realized by decoding surface electromyography (sEMG) signals from users. In terms of signal processing, a sample entropy-based segmentation algorithm, normalized processing and vector quantization, and hidden Markov model (HMM), are proposed. The from-left-to-right method is used to determine the initial HMM parameters. The continuous inputting of letters and characters is realized with the help of AEVIOUS virtual sliding keyboard, and the average online recognition accuracy on four subjects is 91.8% for 10 numbers and 87.6% for 26 letters.

An optimal selection method of servo motor and reduction ratio for high-speed parallel robots is proposed in this paper. In order to achieve the highest speed, relationships between the maximum velocity in a given motion and all the candidate motors/reduction ratio combinations are presented in a graph. It is easy to find the optimal choice and identify the highest speed through this graph. An index is proposed to evaluate the motors’ continuous workability which is the major determining factor of the highest speed. Finally, the method is used in the servo motor and reduction ratio selection of a high-speed parallel robot. The work presented in this paper is very helpful to the development of this robot.

Collision safety is very important when it comes to human-robot collaboration. However, most of the industrial robots are equipped with large capacity motors for high performance, which on the contrary, increases risk of injuries caused by collision between humans and robots. For this reason, collaboration with common industrial robots has been restricted unless the robots are isolated from workers. To address this problem, in this study we propose a safe collaborative robot arm equipped with the counterbalance mechanism (CBM) and sensorless collision detection method. Furthermore, the sensorless collision detection scheme includes the reaction scheme in case of collisions without the use of any expensive sensors. The performance of the proposed CBM and collision detection method was evaluated based on simulations and experiments, respectively. The simulation and experimental results show that it is possible to operate the robot even with only small capacity motors while maintaining its performance, and collision can be reliably detected without any extra sensors for any type of robot manipulator.

This paper presents a dedicated off-line programming system for robotic cutting in pipe processing based on Visual Studio 2010 and OpenGL. The off-line programming system can calculate the space trajectory and generate the programs for a 6-DOF industrial robot automatically. It has a convenient and friendly interactive interface, and the process of cutting can be seen intuitively. Robot programs generated by using the developed system have been tested on a robotic cutting system, and the experimental results are provided, which demonstrate the practicability of the off-line programming system.

Design of the hollow modular joint plays an important role in modern robot layout, fix and wiring. This paper designed a novel hollow modular joint which can meet the requirement of the slave hand of a Minimally Invasive Surgical robot. The mechanical and control design is illustrated sequentially. The torque sensor and its optimization are given. To analyze the designed module, the simulation of the redundant slave robot which is composed by the designed joint in 7 DOF is presented. The results of analyses show that the designed hollow modular joint is valid.

Abstract

In this paper, a new type of 5-DOF hybrid parallel robot is introduced. The architecture of the mechanism is comprised of two rotations attached to the movable platform of a traditional 3-DOF Delta robot. The 5-DOF hybrid parallel robot has larger working space than Delta Robot and can be widely used in many fields. The forward and inverse kinematics of the hybrid robot was established. Then, the experiment was taken to prove its feasibility.

This paper proposes a model prediction control of two wheeled mobile manipulator. The combination of manipulator and two wheeled mobile robot (T-WMR) has been the focus of research in recent year. However, the dynamic analysis is difficult for a multi-DOF mobile manipulator. Moreover, the dynamic equations are complex and will cost great amount of computation time. Real time control is difficult to be realized. Therefore, the system is un-integrated and controlled independently. The effectiveness of the proposed control method has been demonstrated through the experiments.

An experiment of kinematic calibration is conducted to improve the positioning accuracy of a 6-PSS parallel manipulator for large components assembly. To ensure the positioning accuracy of the manipulator with a higher level, the end effector’s pose and position data, which are obtained by the laser tracker, are respectively used in the calibration. The accuracy improvements are compared between the usages of poses and position data and the result shows that though the measurement device has the capability of obtaining the six dimensional poses of the end effector, it is still better to use the position data rather than the pose data to do the kinematic calibration, especially for some large scaled components positioning, for the orientation error may make a bigger influence than the position error in the positioning of the end effector.

In this paper, a contact-less power charger for robot applications is studied and developed. Contact-less charging can be achieved by a separable magnetic-transformer design. The transformer primary core is in the charger unit and the secondary core is in the robots. The transformer air-gap is equal to the distance between these two parts. By theoretical analysis, software simulations and circuit implementation, the relationships between the transformer’s coupling coefficient and the core geometry and gap are formulated. In addition, a high-efficiency circuit topology for the studied contact-less charger is fulfilled. It is anticipated that the research results of this paper can contribute to the development of the contact-less charging techniques for robot systems.

To pursue high-speed motion control for robots, applications of servo motors with appropriate controllers may render fast responses with satisfactory precision. However, the current coupling effect on the servo motor control unavoidably downgrades stability and contouring precision as the operation speed increases in practice. This paper proposes a control structure simply by adding a decoupling gain

Kc

to significantly improve stability of motor control systems under high-speed operations. Simulation results on the joints of a Delta robot indicate that the present decoupling control structure effectively leads to satisfactory performance and precision under high-speed operations.

Rugged terrain robot designs are important for field robotics missions. A number of commercial platforms are available, however, at an impressive price. In this paper, we describe the hardware and software component of a low-cost wheeled rugged-terrain robot. The robot is based on an electric children quad bike and is modified to be driven by wire. In terms of climbing properties, operation time and payload it can compete with some of the commercially available platforms, but at a far lower price.

Specific to the structured operating environment of multi-rows pipeline, the mode of alternated crawling motion with double claws is employed for kinematic and motion planning. Firstly, designing a 5-DOF robot prototype, and building its kinematical model by employing DENAVIT – HARTENBERG Method for forward and inverse kinematical validation; secondly, decomposing the motions in the structured pipeline environment into three typical motions: single pipe crawling, switch between pipes and between rows of pipes, then determining key gestures in the process of crawling, and then performing interpolation of joint space trajectory; finally, performing kinematics simulation analysis based on Adams simulation environment software. The result indicates that: the robot prototype can fulfill the crawling motion trajectory along pipe, between pipes, and between rows of pipes; the joint will sustain more force under over-restrained conditions.

Consider the two characteristics: (1) Simultaneous localization and mapping (SLAM) is a popular algorithm for autonomous underwater vehicle, but visual SLAM is significantly influenced by weak illumination. (2) Geomagnetism-aided navigation and gravity-aided navigation are equally important methods in the field of vehicle navigation, but both are affected heavily by time-varying noises and terrain fluctuations. However, magnetic gradient vector can avoid the influence of time-varying noises, and is less affected by terrain fluctuations. To this end, we propose an adaptive SLAM-based magnetic gradient aided navigation with the following advantages: (1) Adaptive SLAM is an efficient way to deal with uncertainty of the measurement model. (2) Magnetic gradient inversion equation is a good alternative to be used as measurement equation in visual SLAM-denied environment. Experimental results show that our proposed method is an effective solution, combining magnetic gradient information with SLAM.

A novel kinodynamic planning framework, which covers path panning, smoothing, tracking and emergency threat managing, is proposed. The framework is proposed based on sampling-based algorithm, which is improved to ensure dynamics feasibility as well as emergency threat management ability by applying Bezier curve and Extending Forbidden respectively. The Bezier curve guarantees both

$$G^{1}$$

and

$$G^{2}$$

continuity to decrease the tracking error of our LQI based tracking controller, where two Bezier curves with different continuity order are discussed. Extending Forbidden is firstly proposed by us to enable generating multiple paths of sampling-based algorithms, thus support on-line switching to avoid emergency threats. Our main contribution is that the proposed framework is a combination of path planning with emergency threat managing, where a time compromised moving obstacle avoiding method is proposed. Results proves the efficiency of the proposed algorithm in generating feasible trajectory for SERVOHELI-40, which not only guarantees kinematic feasible of avoiding obstacles, but also can ensure dynamics feasibility.

In this paper, a human-machine shared control strategy is developed for the navigation of a wheelchair. The shared controller switches between a brain-machine control mode and an autonomous control mode. In the brain-machine control mode, a novel brain-machine interface (BMI) using only two command signals produced by steady state visual evoked potentials (SSVEP) instead of traditional four-direction command signals is developed. These two brain signals are involved to generate a polar polynomial trajectory (PPT), which is continuous in curvature without violating dynamic constraints of the wheelchair. In the autonomous control mode, the synthesis of angle-based potential field (APF) and vision-based simultaneous localization and map-building (SLAM) technique is proposed to guide the robot navigating in environments where obstacles exist. Experimental studies have been carried out with a number of volunteers and the effectiveness of the proposed shared control scheme has been verified.

A dynamic threat and disturbance rejected path reshaping method is proposed. The method is based upon parametric Bezier curve called Local Optimal Reshaping(LOR), which is easy to adjust the reference velocity for navigation. Before implementation, only minimal safe margin and maximum curvature are needed. The method also purposefully biases the reshaping region of each node, thus, it is computational efficient with easy implementation. Three parts are included in the whole path planner, which are kinematic path planner, disturbance rejector with path smoother, and dynamic threat avoided planner, respectively. LOR acts as main effector in disturbance rejector and dynamic threat avoided planner. Comparative simulations are provided in this paper, and results show that our method has good performance in tackling disturbance and dynamic threats.

This paper presents a developmental approach to create a mobile robot’s reaching ability. The approach consists of two stages. In the first stage, the mobile robot cannot move, so that the robot develops the reaching ability in the static environment. In the second stage, the robot learns to control its moving ability. By combining the two abilities, the robot successfully obtains the ability to reach objects that are far away from the robot. The two abilities are implemented by two artificial neural networks. A developmental constraint driven mechanism is applied to the robot system, so that, the robot adapts to the environment and completes the tasks in dynamic environment step by step. The experiments and simulations demonstrate that the robotic system, by imitating the process of the human development, gradually and autonomously obtains the reaching ability.

In order to guarantee high-quality original images and achieve automatic control of lighting intensity in the motion of inspection robot following the planned trajectory, this paper proposes a time and light intensity optimal trajectory planning method that not only considering kinematics constraints traditionally, but also take task demands of vision inspection into consideration. We study the influence of light intensity on image quality, and put forward image quality assessment metrics related to light intensity, model the mathematical relationship among light intensity, image quality assessment metrics and camera observation distance. And then optimization model based that mathematical model for time and light intensity optimal trajectory planning is presented. Experiments demonstrate the feasibility and efficiency of this method.

In this paper the solutions of forward and inverse kinematics problem for Adept Six 300 robot are presented. A matrix method was used in computations. Presented algorithm was used in trajectory planning. Joint values are chosen from a given set of solutions. A graph of solutions is made. Nodes represent the solutions of inverse kinematics problem and arcs’ weight are the cost of transition between two configurations. Two criteria are proposed in order to obtain smooth motion with desired properties. Algorithm based on derived formulas and logical conditions has been implemented in Adept SmartController CX. A comparison between manufacturer’s and proposed algorithm is presented. Proposed solution of inverse kinematics problem allows for a smooth passage through singular points.

To enhance the robustness in tolerance for ZMP design imperfection and for uncertain external disturbances in online biped gait generation, a modified preview control strategy with angle coefficient adjustment is presented. First, an offline walking pattern generator based on ZMP stability criterion and an online one using ZMP preview control with an observer integrated are built. Then the concept of angle coefficient is proposed and the method for angle coefficient real-time adjustment is presented. To evaluate the effectiveness of proposed method, simulation experiments are conducted with a 10 DOFs prototype. The results show that, using the proposed method, the robot prototype is capable of reducing the center of mass (COM) trajectory tracking error; the robot prototype is capable of reducing walking yaw in spite of ZMP design imperfection and modeling error; the robot prototype is capable of walking stably through uneven path surfaces.

A flexible ankle joint for biped walking robots is proposed to investigate the influence of joint stiffness on motor’s peak torque and energy consumption of the sagittal plane motion during the single support phase. Firstly, an improved model of the inverted pendulum is established, which is the theoretical foundation of the flexible ankle joint. Then the analysis of the analytic method of flexible joint is presented based on the improved model of the inverted pendulum. Finally, dynamic simulations of the flexible joint are performed to examine the correctness of analytic method. The results show that the flexible joint can reduce the joint motor’s peak torque and energy consumption. Furthermore, there is an optimal joint stiffness of the flexible system, which can minimum peak torque with reduction of 45.99% and energy consumption with reduction of 51.65%.

This paper addresses the problem of properly placing a given task in the manipulator workspace by a heuristic and numeric approach. Thus, the task is placed relatively to the manipulator for each element of the discretized workspace and the required joint torques are determined. The results are are by a torque-based optimization criterion. The modularity of this approach ensures general applicability on various systems and tasks while the high computational effort is treated by GPU parallelization. The method is presented for a given 6DOF manipulator and a highly dynamic trajectory. The resulting interactive map of the manipulator workspace gives an overview of the task dependent dynamic performance, detailed evaluation of certain solutions will show the dexterity of the proposed approach.

Currently, an automated surface treatment or finishing (e.g., abrasive blasting, cleaning or painting) is performed in two consecutive steps: first processing by tool, second quality evaluation by sensor. Often, a finished surface has defects like areas not properly processed. This is caused by path inaccuracy or errors in tool deployment. The defected areas can be detected only during a subsequent quality evaluation. As a result, a complete surface reprocessing is required which is costly and time-consuming. We propose a new approach that is a combination of surface treatment and quality evaluation processes in a single deployment. In our approach, an initial coverage path for surface treatment by a tool is given or calculated using a state-of-the-art coverage path planning algorithm. In fact, we extend an off-line generated initial coverage path with an ability to react to sensor-based defect recognitions during surface processing by a tool.

The improvement of stability for dynamic biped walking is motivated by the potential use of humanoid robot in complex environment. This paper focuses on the effect of ground compliance on the periodic stability of dynamic walking. Firstly, the compliant ground is equivalent to a spring-damper system, and the coupling dynamics model of robot-ground system is modelled as a rigid planar kinematic chains in contact with a stiffness-damping system. Based on it, two differential equations are obtained to describe the dynamic walking process in one cycle which is separated into a swing phase followed by an impact phase. Subsequently, a stable gait is planned under rigid assumption and transplanted to the walking under compliant ground condition. The effect of ground compliance on the gait during the swing phases is analyzed under 15 kinds of ground conditions, respectively and effect on it during impact phase is analyzed under 12 kinds of ground conditions. Finally, the periodic stability under the effect of ground compliance is analyzed with 6 kinds of ground conditions, and the effect of ground compliance on the periodic stability of dynamic walking is concluded.

This paper presents an adaptive compliant multi-finger grasp approach control strategy based on based on a new interpretation of the virtual spatial spring framework, to improve the grasp performance for target objects with position errors. An n-finger virtual spatial spring frame is proposed to achieve the adaptive compliant grasp control. Two-finger grasp control based on a single virtual spring is tackled, and then extended to multi-finger grasp control. Virtual springs for self-collision avoidance among digits are constructed to form the complete adaptive compliant grasp control law. With the virtual-spring based adaptive compliant grasp approach control strategy, the first robot finger to experience unexpected impact remains in contact with the object, while the rest of the fingers are continuously, adaptively driven toward re-adjusted grasping positions by the virtual springs without the need for on-line replanning. Experimental results demonstrate effectiveness of the virtual-spring based grasp controller, and significantly larger position errors of the target object can be accommodated with the proposed adaptive compliant grasp control strategy.

This work presents a synchronized cross-coupled control strategy for the coordination of multi-manipulators system. In the paper, the model-free synchronization error definition manner for different manipulators is introduced in detail. In addition, the joint actuator load of the system is translated and compensated to further improve the synchronization performance of the system. Taking a typical mirror milling system as the research object, the system model and control model are established, the simulation result shows that the proposed strategy can effectively improve the coordination for multi-manipulators system.

A novel redundantly actuated parallel manipulator is developed in this paper. With the aid of redundant actuation, this manipulator can conquer Type II singularities within the workspace of the traditional planar 5R parallel manipulator. To investigate the dynamics, the D’Alembert principle is utilized to establish the dynamic model of the manipulator. Based on the method of minimizing the Euclidian 2-norm, the driving torques of active joints can be obtained, and then the constraint forces of all joints can also be obtained. Through inverse dynamic simulation, it is demonstrated that the redundantly actuated parallel manipulator has better dynamic performance in the nearby zone of Type II singularities than the traditional planar 5R parallel manipulator.

This paper focuses on proposing a real-time path-smoothing method for DELTA parallel robots, which utilizes the 5-th Bézier curve to blend the adjacent linear segments. By using the proposed method, a smoothing trajectory with the continuities of second-order derivative and curvature is generated. Meanwhile, an S-shape acceleration/deceleration (ACC/DEC) process and a predictor-corrector interpolator are employed for reducing the mechanism vibration and velocity fluctuation. Therefore, smoothing transition between vertical and horizontal segments of the operation path can be guaranteed. Experiments are carried out on the self-designed DELTA parallel robot and the results verify the feasibility of the proposed real-time path-smoothing method.

This paper presents a simulation framework for using machine learning techniques to determine robust robotic motions for handling deformable objects. The main focus is on applications in the meat sector, which mainly handle three-dimensional objects. In order to optimize the robotic handling, the robot motions have been parametrized in terms of grasp points, robot trajectory and robot speed. The motions are evaluated based on a dynamic simulation environment for robotic control of deformable objects. The evaluation indicates certain parameter setups, which produce robust motions in the simulated environment, and based on a visual analysis indicate satisfactory solutions for a real world system.

This paper presents a new novel SINGLE-PORT access surgery (SPS) robot system. This surgical robot is composed of a surgical slave robot with flexible surgical instruments and an ergonomic master device with an image guided system. This surgical slave robot has a six-degrees-of-freedom (6-DOF) guide tube, two 7-DOF surgical tools, a 3-DOF stereo-endoscope and a 5-DOF slave arm high. The master device has a 14-DOF ergonomic instrument controller and a three-dimentional image guided system. Therefore, the operator can approach surgical instruments to the target with various poses through the master console. The experimental results for surgical operations shows the feasibility of this robot in the field of robotic surgery through single-port.

Simple expert systems are presented that will allow more people to use powered wheelchairs. The systems interpret hand tremor and provide joystick position signals. Signals are mixed with ultrasonic sensor data to identify potentially hazardous situations and assist users to find a safe course. Results are discussed from a series of timed tasks completed by users using a joystick. They suggest that the amount of sensor support should be varied depending on circumstances and skill. Drivers completed progressively more complicated courses both with and with-out sensors and the most recently published systems are used to compare results. The new expert systems consistently out-performed the most recently published systems.

Cardiac ablation therapy is an effective minimally invasive treatment of cardiac arrhythmias. The procedure is delicate and complex in nature requiring particular sections of the heart to be ablated. When operated by an experienced electrophysiologist, the procedure normally takes 2 to 3 hours. It is often conducted under fluoroscopic X-ray guidance and longer procedure times can increase the radiation burden of both the operator and patient. Earlier, we had proposed a robot-assisted tendon-guided catheter that can be navigated using radiation-free MRI imaging. In this paper, we propose a tension-feedback mechanism that provides vital control over its guiding tendons. We describe how it can be achieved using tension sensing and demonstrate using experiments and finite element simulations that feedback based on accurate tension sensing is plausible.

This paper presents a methodology for learning and adaptation of a 3-PRS parallel robot skills for ankle rehabilitation. Passive exercises have been designed to train dorsi/plantar flexion, inversion/eversion ankle movements. During exercises, forces may be high because patient cannot follow the desired trajectory. While small errors in the desired trajectory can cause important deviations in the desired forces, pure position control is inappropriate for tasks that require physical contact with the environment. The proposed algorithm takes as input the reference trajectory and force profile, then adapts the robot movement by introducing small offsets to the reference trajectory so that the resulting forces exerted by the patient match the reference profile. The learning procedure is based on Dynamic Movement Primitives (DMPs).

Nowadays wearable rehabilitation exoskeletons are one of the most studied gait rehabilitation tool from a technological point of view. Current devices use prerecorded healthy gait patterns. This leads to potentially non-natural imposed gait patterns, and to solve this issue, we propose the use of regression-based methods to reconstruct speed dependent angular trajectories. Results suggest that the proposed method can lead to a more natural gait. Consequently, a naive user may more easily learn to walk under the presence of a robot guidance.

The transfer of harmful aquatic organisms and pathogens in the ballast water of ships has a significant global impact on coastal regions. The effects have made ballast management systems extremely important for protecting marine environments. Recently, to solve this issue, a number of technologies have been developed and commercialized. Most of them are barely used independently, but there are several combined methods to treat ballast water. The aim of this study is to suggest the best method of ballast water sterilization using a high-voltage pulsed arc discharge and a vision monitoring system.

Traditional stack systems for petri dish dispenser are complex in both structure and control systems, since they are driven by four or more motors. A novel coaxial stack system, with functions of extracting, opening, transferring, closing and inserting of petri dishes, is designed in this paper. It is driven by three motors, resulting in a simple control system. This coaxial stack system consists of a stacking frame, a tray, a step plate, a carrier and a piston. The stacking frame and the carrier are rotating coaxially leading to a smaller space occupation. Mechanical analysis and simulation of the key part of the stack system are described in details. The results of theoretical analyses and actual experiment show that the proposed coaxial stack system can work very well.

Hydrogel materials and forming processes have a significant impact on the performance of tissue engineering scaffolds and cell 3D printing. Scaffolds supports cell adhesion as a temporary extracellular matrix, which plays a key role in tissue engineering and regenerative medicine. Hydrogel 3D structure bio-fabrication was an important step towards cell 3D printing. A precision extrusion nozzle based on ball screw transmission was developed for porous hydrogel fabrication. The whole bio-fabrication temperature was controlled between 4-14°C which eliminates ice exactly. Three-dimensional structure was formed based on the temperature-sensitive gelling properties of gelatin. Hereby, the influence of hydrogel concentration, extrusion speed and scanning speed, printing temperature, scanning spacing and the heights of layer were analyzed in depth. Hydrogel scaffolds were fabricated pore network in the gelatin 10% and the sodium alginate 2%. The processing parameters of this paper can be directly applied to hybrid printing of cell and hydrogel material.

Laser bonding is one of important technology for devices Micro-interconnection and encapsulation. Laser has a characteristic of focusing, which makes the temperature of irradiate region dramatically increase in a short time, while the entire sample keeps a low temperature. Therefore, laser bonding has broad application prospects in micro-welding field. However, the welding path planning and positioning for micro-area welding need to overcome the shortcoming of low accuracy via using positioning blocks during traditional welding manner. This paper develops a visual identification and location system using a CCD (charge coupled device) camera. Based on the figure captured by CCD camera, an image processing system is used to extract the position and the size of the solder by PC software LabVIEW. Combined the collected information with the process requirements, the process planning system specifies the servo motor operating parameters and laser power parameters. This system can locate and identify arbitrary rectangular frame in the CCD field of view instead of pre-setting the size and position of solder. What’s more, based on the planned parameter, the switch of laser and the laser power value can adjust to laser path, making the laser processing more flexible.

The characteristic of high frequency ultrasonic vibrations of an ultrasonic dental drill plays a crucial role in performance of superior precision bone surgery. This paper proposes an attempt to enhance its vibration performance based on the composite horn consisting of an ultrasonic horn and working tip. Finite element analysis has been carried out to investigate dynamic analysis of different structures of composite horns including vibration modeling and harmonic response analysis. Natural frequency of the ultrasonic horn is calculated within the range of excitation frequency of ultrasonic horns in the vibration modeling, while the combined displacement and mode shape of the composite horn are obtained in the harmonic response analysis. The dynamic analysis confirms that the natural frequency and the resonance frequency of the composite horn can be adjusted by designing the small end length of the ultrasonic horn, and suitable mode shape of the composite horn can be obtained by changing the shank angle of working tip. The research finding provides theoretical principles for designing an ultrasonic dental drill in dental implantology.

Feedrate scheduling is crucial for the CNC machine tools to achieve high surface quality and machining efficiency. In this paper, an adaptive feedrate scheduling method is proposed to minimize the machining time for five-axis machining. First, a feedrate optimization model with constraints of chord error, maximum velocity, acceleration and jerk of five drive axes is presented. To guarantee machining stability in the process of manufacturing with high feedrate, the maximum tangential feedrate, acceleration and jerk are also considered. Then, the chord error and acceleration constraints are transferred to linear inequalities by discretizing the tool path. Meanwhile, the jerk constraints are simplified with conservative linear constraints. Finally, the feedrate scheduling problem is transferred to a line-ar programming problem, which can be solved efficiently. Simulation and experiments are conducted to verify the feasibility and efficiency of the proposed method.

Five-axis machine tools are widely utilized in the manufacturing of the parts with complex geometries, and contour error is the significant measurement for the accuracy of the final production. Generally, a high accuracy estimation of the contour error is very important since it is the premise of the high quality contouring control. Based on the foot point evaluated by the arc length between the foot point and the reference point nearest to the actual position, a high accuracy real-time contour error estimation method is proposed in this paper, and then the control law is designed to achieve the specified contouring performance. Experimental results show that the contour errors calculated by the proposed method are very close to the real ones, which demonstrates the effectiveness of the proposed method.

In this paper, a levitating nano-positioning stage with arc-edged permanent magnet Halbach arrays is presented. Construction of the nano-positioning stage is given, and the components of the stage are designed. For the key actuation unit, an accurate model of the arc-edged permanent magnet Halbach actuator is developed using the Fourier series expansion, which is validated by the Maxwell software. Based on the model, optimization is made to the arc-edged permanent magnet Halbach array. The optimization results show that the arc-edged permanent magnet Halbach actuator can produce a very small thrust ripple with a large average thrust to guarantee the high precision of the positioning stage.

The clamping length of holder-tool often changes with the machining conditions and is the main factor affecting the stiffness and damping values of holder-tool interface. To avoid the need for repeating the measurements, the previous studies identify the joint stiffness-damping parameters based on genetic algorithm and calculate the joint parameters per unit area of holder-tool interface for response prediction. This paper proposes a simplified method to identify the joint parameters by simplifying the continuous contact interface to contact points in side of the clamping part. The equivalent joint stiffness-damping parameters are calculated by an inverse calculation method along the whole frequency band, which is more efficient than the genetic algorithm. The responses of holder-tool assembly with different clamping length are predicted based on the assumption of linear relationship between the stiffness-damping values and clamping length. At last, the experiment cases were carried out to verify the effective of the simplified method.

This paper focuses on proposing a tangent prediction NURBS interpolator (TP) method for reducing the machining time and improving the surface quality, which adopts predicted tangent vectors to constrain the curve shape and suppress the curve fluctuation. Although the traditional NURBS path interpolation methods have the advantageous in flexibility and can be easily implemented, those methods tend to curve distortion if insufficient interpolation points or constraints are given, which will extend the curve length and influence the accuracy of the constructed module. In this paper, the geometric module constructed by the TP method can not only ensure a smooth transition between curve sections but also decrease the curve distortion. With the proposed method, the curvature radii at some local areas may increase, which means the chord error will increase accordingly. To minimize the chord error, a curve subsection process along with a 7-segment S-shape acceleration/deceleration (ACC/DEC) process are adopted here. Meanwhile, the predictor-corrector interpolator method is applied to reduce the velocity variation and improve the accuracy of the control system. Experiments are carried out to demonstrate the ability of the method by virtue of reducing the machining time and improving the surface quality.

During the turning process, unstable self-excited vibration known as regenerative chatter adversely affects tool life, workpiece quality and productivity efficiency. Duo to the low-rigidity characteristic, workpiece chatter usually occurs rather than tool in turning of the low-rigidity workpiece. In this work, instead of improving workpiece rigidity by auxiliary sustaining device, an active control methodology is proposed to suppress chatter such that the area of stable cutting is increased and a higher productivity can be obtained. The methodology developed in this paper is based on dynamic system matching of tool and workpiece. Hereto, a fast tool servo (FTS) system is designed to realize system matching, and then the ability to suppress regenerative chatter is evaluated by turning a low-rigidity slender rod. The cutting experiment results demonstrate the validation of the proposed methodology in turning of the low-rigidity workpiece.

The assemble precision of thin-walled structures are very important in engineering, which affects the performance of the mechanism. Large rotations are easy to occur in these structures during assembling process due to their geometric features. So it is necessary to establish a precise model of the thin-walled structure to predict its deformation. In this study, the absolute nodal coordinate formulation (ANCF) method is used to describe initial curved thin-walled structure and to study the compatibility of deformation of thin-walled structures during assembling process. A four-node shell element of ANCF with 48DOFs is employed to discrete the thin-walled structure. The vector of the element elastic forces and the stiffness matrix are derived based on continuum mechanics. Static deformation tests of a cantilever plate and the modal analysis of free square plate are used to validate the formulations for the plate element of ANCF. The statics equilibrium equations are deduced and the compatibility of deformation of two thin-walled structures is investigated.

The nano-positioning system with long stroke is the key component of the micrometer and nanometer manufacturing and measurement devices, such as atomic force microscope (AFM), laser direct writing, nano-machining, lithography etc. In this paper a novel linear magnet-driven actuator is proposed. The Halbach magnetic array with air bearings is adopted as the mover. The winding of stator is ironless structure. The actuator has a nanometer scale positioning resolution. The designed stroke is 50 millimeter. The high-accuracy optical incremental encoder and subdividing system are employed to measure the motion of the mover, facilitating real-time feedback control. The hardware-in-loop simulation system of the actuator is set up based on XPC-target module in Matlab/Simulink toolbox. The parameter identification and Simulink control model of the actuator are implemented. Finally, the trial test and analysis are carried out. The results show the presented actuator could be operated with the positioning resolution of 9.5 nm root mean square (RMS).

In machining applications, active magnetic bearings (AMBs) have great potential to improve efficiency, reduce costs, and enhance product quality, due to its high force capacity, high speed capability, and ability to monitor states and employ active controls. This paper describes structure, rotor dynamic modeling and control of a milling spindle with AMBs. The rotor dynamic model is created using FEM according to the Timoshenko beam theory. Before applying the spindle to suppress milling chatter, we develop an

$$\mu $$

-synthesis design framework to stabilize the unload system considering the uncertainty in spindle speed. Balance truncation procedure is implemented for the controller order reduction. Finally, a 15

th

order controller which ensures performance requirements and robust stability, is obtained. The spindle with the controller has speed range of 0–40000 rpm.

In aerospace industry, heat-resistant-super-alloys (HRSA) are widely used for its high strength, high hardness and low thermal diffusivity at high temperature. Flank milling is performed by employing the side of a cutter to touch the desired workpiece surface. Compared to traditional point milling, flank milling has attracted interest because of its high material removal rate and no scallops on the workpiece for single passage flank milling. Experimental investigation of the cutting performance in flank milling of HRSA is a preliminary step to the practical usage of HRSA, although this issue is absent to date. This paper firstly reviews the state of the art on the cutting forces and surface integrity in HRSA milling process. Then the experimental studies of cutting forces variation and surface integrity in flank milling of HRSA with PVD coated cemented carbide tools are presented. The cutting tests are performed at various cutting speeds (40–120 m/min), cutting depths (9–15 mm) and feed rates (0.03–0.09 mm/rev) in flank milling under lubricant condition. Additional measurements (surface roughness, residual stresses, microstructure and micro-hardness) are performed with the sample of the machined surface in addition to the cutting forces. The feed rate is the most influential parameter with surface roughness in flank milling operation. The tensile residual stresses are observed on the machined surface and a work hardening layer arises beneath the machined surface. The results show that the lower cutting forces and smaller micro-hardness generated with the higher cutting speed. The detailed analysis of the relationship between cutting parameters and surface micro-hardness is given in this paper.

This paper presents the prediction of cutting forces in micro end-milling process based on the orthogonal cutting finite element simulation, which includes the exact tool trajectory with run-out, edge radius, rake angle, tool-chip contact and material mechanical and physical properties. The Johnson-Cook constitutive model parameters and sliding friction coefficient on the tool-chip interface are determined by the orthogonal cutting experiments. The finite element simulations of the micro orthogonal cutting process are carried out to determine the relations between tangential and feed cutting forces and uncut chip thickness by using the AdvantEdge FEM software. The effect of the run-out on the tool trajectory is considered to determine the exact uncut chip thickness. The cutting forces model is validated by micro slot end-milling experiments carried out on a three-axis ultra-precision machine. The predicted cutting forces are in good agreement with the experimental results.

Most vibro-impact oscillators in engineering applications appear to include nonlinear stiffness or damping, but little attention has been paid to this kind of oscillator. Thus, in present paper, a physical model for an undamped and periodically forced Duffing oscillator with a constraint which leads to motions impacts was analyzed. Computational method was used to solve the nonlinear governing equations. Rich dynamical behaviors including periodic motion, chaotic motion, chattering and grazing were observed in this simple system. Influence of non-dimensional system parameters including the nonlinear stiffness coefficient

β

, the forced frequency Ω, the clearance Δ on motion character of the system were also discussed through corresponding bifurcation diagrams. It is supposed that: (a) chattering appears when Δ less than certain threshold. (b) chaotic motion arises when

β

larger than certain threshold. (c) grazing bifurcation occurs when

r

larger than certain threshold.

This paper presents a high-speed tracking control approach for third-order piezo-actuated nanopositioning stages, which extends the vibration control strategies tailored for damping the resonant modes of second-order systems (SOSs) to third-order systems (TOSs). The extension consists of three steps: i) decomposing the TOS into a SOS and a first-order system (FOS); ii) designing the vibration controller for the SOS; iii) extending the vibration controller to the TOS by cascading the controller with the inversion of the FOS. To illustrate the effectiveness of the proposed approach, the positive position feedback (PPF) controller cascaded with dc-gain inversion of FOS is designed. The extended PPF controller is adopted in the inner feedback loop to damp the resonant mode of the system. Then, in the outer loop, a high-gain proportional-integral (PI) controller is utilized to minimize the tracking errors due to disturbances and modeling uncertainties. Experimental results on a piezo-actuated nanopositioning stage demonstrate that the proposed control approach achieves high-speed tracking by improving the control bandwidth from 80 Hz (with PI controller) to 322 Hz.

The thin-walled structures always tend to distort and deform during machining due to its low rigidity, which will influence the tolerance characteristics of the final product. Thus, to control machining accuracy of is very challenging. This paper focuses on a solution to this problem, an information-localizing method for machining large thin-wall structures is proposed. Further, a fiducial correction algorithm is used to compensate the machining deformation error. Finally, the experiments were carried out, and the results indicated that the proposed method can improve the machining accuracy.