Showing papers by "Zeng-Guang Hou published in 2013"

Distributed consensus control for multi-agent systems using terminal sliding mode and Chebyshev neural networks

Abstract: This paper investigates the problem of consensus tracking control for second-order multi-agent systems in the presence of uncertain dynamics and bounded external disturbances. The communication ?ow among neighbor agents is described by an undirected connected graph. A fast terminal sliding manifold based on lumped state errors that include absolute and relative state errors is proposed, and then a distributed finite-time consensus tracking controller is developed by using terminal sliding mode and Chebyshev neural networks. In the proposed control scheme, Chebyshev neural networks are used as universal approximators to learn unknown nonlinear functions in the agent dynamics online, and a robust control term using the hyperbolic tangent function is applied to counteract neural-network approximation errors and external disturbances, which makes the proposed controller be continuous and hence chattering-free. Meanwhile, a smooth projection algorithm is employed to guarantee that estimated parameters remain within some known bounded sets. Furthermore, the proposed control scheme for each agent only employs the information of its neighbor agents and guarantees a group of agents to track a time-varying reference trajectory even when the reference signals are available to only a subset of the group members. Most importantly, finite-time stability in both the reaching phase and the sliding phase is guaranteed by a Lyapunov-based approach. Finally, numerical simulations are presented to demonstrate the performance of the proposed controller and show that the proposed controller exceeds to a linear hyperplane-based sliding mode controller. Copyright (C) 2011 John Wiley & Sons, Ltd.

Improved Rao-Blackwellized Particle Filter by Particle Swarm Optimization

Abstract: The Rao-Blackwellized particle filter (RBPF) algorithm usually has better performance than the traditional particle filter (PF) by utilizing conditional dependency relationships between parts of the state variables to estimate. By doing so, RBPF could not only improve the estimation precision but also reduce the overall computational complexity. However, the computational burden is still too high for many real-time applications. To improve the efficiency of RBPF, the particle swarm optimization (PSO) is applied to drive all the particles to the regions where their likelihoods are high in the nonlinear area. So only a small number of particles are needed to participate in the required computation. The experimental results demonstrate that this novel algorithm is more efficient than the standard RBPF.

An enhanced dual-finger robotic Hand for Catheter manipulating in vascular intervention: A preliminary study

Abstract: Cardiovascular disease is a class of diseases that involve the heart and blood vessels, which is the leading cause of deaths globally. Vascular interventional surgery is an effective treatment. As belongs to minimally invasive surgery (MIS), it is more popular due to the advantages of smaller injury, faster recovery and higher accuracy rate. However, it needs up to 10 years to train a skilled surgeon for minimally invasive cardiovascular surgeries. The surgeon cannot avoid receiving the excessive dose of X-ray due to long time operation in daily work. A dual-finger robotic hand (DRH) was introduced to assist the surgeon in manipulating the catheter/guide wire. As a surgical device, DRH is aiming at simple mechanism, ease to use and convenient sterilization. It was carefully investigated how the surgeon manipulates the catheter/guide wire. The bionic thumb and forefinger were designed to imitate human's. Compared to human's, the two bionic fingers are enhanced due to that they can advance the catheter/guide wire without moving the whole hand. The DRH mechanism was carefully designed. A console was also designed for the surgeon to manipulate DRH. After the DRH control was done, the minimum robotic cardiovascular interventional system was formed. Experimental results have validated the feasibility of DRH and the robotic system. Future work regarding DRH is also discussed.

NeuCubeRehab: A Pilot Study for EEG Classification in Rehabilitation Practice Based on Spiking Neural Networks

Abstract: One of the most important issues among active rehabilitation technique is how to extract the voluntary intention of patient through bio-signals, especially EEG signal. This pilot study investigates the feasibility of utilizing a 3D spiking neural networks-based architecture named NeuCube for EEG data classification in the rehabilitation practice. In this paper, the architecture of the NeuCube is designed and a Functional Electrical Stimulation (FES) rehabilitation scenario is introduced which requires accurate classification of EEG signal to achieve active FES control. Three classes of EEG signals corresponding to three imaginary wrist motions are collected and classified. The NeuCube architecture provides promising classification results, which demonstrates our proposed method is capable of extracting the voluntary intention in the rehabilitation practice.

Combined use of sEMG and accelerometer in hand motion classification considering forearm rotation

Abstract: Hand motion classification using surface electromyography (sEMG) has been widely studied for its applications in upper-limb prosthesis and human-machine interface etc. Pattern-recognition based control methods have many advantages, and the reported classification accuracy can meet the requirements of practical applications. However, the pattern instability of sEMG in actual use limited their real implementations, and limb position variations may be one of the potential factors. In this paper, we give a pilot study of the reverse effect of forearm rotations on hand motion classification, and the results show that the forearm rotations can substantially degrade the classifier's performance: the average intra-position error is only 2.4%, but the average interposition classification error is as high as 44.0%. To solve this problem, we use an extra accelerometer to estimate the forearm rotation angles, and the best combination of sEMG data and accelerometer outputs can reduce the average classification error to 3.3%.

Necessary and sufficient conditions for solving leader-following problem of multi-agent systems with communication noises

Abstract: The leader-following problem of first-order integral multi-agent systems with communication noises is investigated in this paper. To attenuate the noise's effect, a positive time-varying gain a(t) is employed in the protocol. It is proved that the proposed protocol can solve the mean square leader-following problem if the following conditions hold: 1) the communication topology graph has a spanning tree; 2) ∫0∞ a(t)dt = ∞; 3) limt→∞ a(t) = 0. The requirements on a(t) are different from most existing papers, where a(t) is required to satisfy that ∫0∞ a(t)dt = ∞ and ∫0∞ a2(t)dt <; ∞. It turns out that ∫0∞ a2(t)dt <; ∞ implies limt→∞ a(t) = 0, if a(t) is uniformly continuous. Therefore this paper relaxes the requirements on a(t) to some extent. In addition, under the mild condition (a(t) is uniformly continuous) these three conditions are necessary as well. Furthermore, if ∫0∞ a2(t)dt <; ∞, the employed protocol is proved to be able to solve the almost sure leader-following problem of first-order integral multi-agent system. Finally, a simulation example is provided to verify the effectiveness of the employed protocols.

A multi-body mass-spring model for virtual reality training simulators based on a robotic guide wire operating system

Abstract: Generally, surgeons of minimally invasive surgery should possess good ability to coordinate their both hands. The manipulation of guide wire is considered a core skill. Obtaining that core skill to perform minimally invasive surgery requires training. In minimally invasive surgery, a surgical robot can assist doctors to position precisely and provide stable operation platform. Therefore, to develop a virtual reality simulators for training purpose based on a robotic guide wire operating system is an important and challenging subject. In this paper, a multi-body mass-spring model for simulating guide wire is presented and evaluated. In order to overcome the disadvantage of using mass-spring approach to model the guide wire, we propose a new collision detection algorithm and a new collision response algorithm. Finally, we test our guide wire with a complex and realistic 3D vascular model, which is selected from computer tomography database of real patients. The result shows that the virtual reality training simulators is effective and promising.

Backward swimming gaits for a carangiform robotic fish

Abstract: This paper focuses on the gaits planning method of the backward swimming for unsymmetrical structure bio-inspired robotic fish. Based on the differences between the anguilliform mode and carangiform mode swimming, a method for searching gaits of backward swimming was proposed to plan the motion of the developed carangiform robotic fish. The body envelope of European eel’s backward swimming was mimicked according to the freely swimming model, which was proposed to analyze the propulsion produced by the undulation of the multi-link tail. Finally, simulations and experiments were conducted to demonstrate the gaits searching method for the bio-inspired carangiform robotic fish.

3D modeling of coronary arteries based on tubular-enhanced CURVES segmented regions for robotic surgical simulation

Abstract: The visualization of the coronary vasculature is of utmost importance in interventional cardiology. Intravascular surgical robots assist the practitioners to perform the complex procedure while protecting them from the tremendous occupational hazards. Robotic surgical simulation aims to provide support for the learners in both efficiency and convenience. The blood vessels especially the coronary arteries with rich details are the key part of the anatomic scenario of the virtual training system. The variations in diameters and directions make the segmentation of the coronary arteries a difficult work. In this paper, a robust and semi-automatic approach for the segmentation of the coronary arteries is developed. The approach is based on the multi-scale tubular enhancement and an improved geodesic active contours model. The demonstrated approach firstly enhances the tubular objects by computing their “vesselness”. Next the edge potential maps are calculated based on the enhanced information. Meanwhile, the initial contours are generated by a modified fast marching method. Then the actual wave fronts evolution extracts the details of the coronary arteries. Finally the visualization model is organized based on the segmentation results by the marching cubes method. This approach has been proved successful for the visualization of the coronary arteries based on the CTA information.

sEMG Based Joint Angle Estimation of Lower Limbs Using LS-SVM

Abstract: In this paper, a new estimation model based on least squares support vector machine (LS-SVM) is proposed to build up the relationship between Surface electromyogram (sEMG) signal and joint angle of the lower limb. The input of the model is 2 channels of preprocessed sEMG signal. The outputs of the model are joint angles of the hip and the knee. sEMG signal is acquired from 7 motion muscles in treadmill exercise. And two channels of them are selected for dynamic angle estimation for their strong correlation with angle data. Angle estimation model is constructed by 2 independent LS-SVM based regression model with radial basis function (RBF). It is trained using part of the sample sets acquired in 10s exercise duration and test by all data. Experimental result shows proposed method has good performance on joint angles estimation based sEMG. Root mean square error (RMSE) of prediction knee and hip joint angles is 3.02° and 2.09° respectively. It provide new human-machine interface for active rehabilitation training of SCI, stroke or neurological injury patients.

An FES-assisted training strategy combined with impedance control for a lower limb rehabilitation robot

Abstract: In order to investigate the feasibility of integrating functional electrical stimulation (FES) with robot-based rehabilitation training, this paper proposes an FES-assisted training strategy combined with impedance control for our self-made exoskeleton lower limb rehabilitation robot. This control strategy is carried out in a leg press task. Through impedance control, an active compliance of the robot is established, and the patient's voluntary effort to accomplish the task is inspired. During the training process, the patient's related muscles are applied with FES which provides an extra assistance to the patient. The intensity of the FES is properly chosen aiming to induce a desired active torque which is proportional to the voluntary effort of the patient. This kind of enhancement serves as a positive feedback which reminds the patient of the correct attempt to fulfill the desired motion. FES control is conducted by a combination of neural network-based feedforward controller and a PD feedback controller. The feasibility of this control strategy has been verified in Matlab.

EMG-based estimation of knee joint angle under functional electrical stimulation using an artificial neural network

Abstract: In this paper, an artificial neural network is proposed to estimate knee joint angle in hybrid activation of knee extension motion, including voluntary muscle contraction and functional electrical stimulation (FES) induced contraction. Voluntary electromyography (EMG) signals of three muscles responsible for knee extension and FES parameter which describe the FES intensity are used as input vector of the neural network, while the estimated knee angle is the output. During the experiment, FES with different combinations of parameters (pulse amplitude and pulse width) was delivered to the rectus femoris muscle of a healthy male subject when the knee was in a periodic extension motion by voluntary muscle contraction. Raw EMG signals of three muscles, parameters of FES as well as the actual knee angle were recorded. Totally, there were 52,233 and 17,420 sampling points corresponding to 261 and 87 seconds used to train and validate the neural network. The result shows the trained network has a satisfactory performance on knee joint angle estimation whose output well follows the curve of actual knee angle. Root mean square error between estimated angle and actual angle is employed to represent the estimation accuracy which is 5.07 degree according to the experimental data.

Crus adjustment system for rehabilitation robot

Abstract: A crus adjustment system for a rehabilitation robot includes a crus extension/contraction unit and a knee joint rotation unit. A motor (15) drives a screw rod (2) to rotate, and accordingly drives a screw rod nut and a connecting frame (4) to cause a dovetail groove guide rail (1) fixed to the connecting frame (4) to move, thus enabling extension/contraction of the crus during rehabilitation training. A pulley (10) drives a torque sensor (25) fixed on a torque sensor seat (24), and accordingly drives the whole knee joint, thus enabling the knee joint to rotate. The system has two degrees of freedom, and has the advantages of high positional precision, good intuitive sense for movement, high overall rigidity, simple structure and easy manufacture.

Consensus tracking of general linear multi-agent systems: Fast sliding-mode algorithms

Abstract: This paper solves distributed consensus tracking problems where the task is to make the multi-agent network, with each agent described by a general linear dynamics, to reach consensus with a leader whose control input is nonzero and not available to any followers. A set of sliding mode surfaces are defined and then fast sliding mode controllers are designed for both reduced order and non-reduced order cases. It is shown that all the trajectories exponentially converge to the sliding mode surfaces in a finite time if the leader has a directed path to at least one of the followers in a strongly connected and detailed balanced directed interaction graph and the leader's control input is bounded. The control Lyapunov function for exponential finite time stability, motivated by the fast terminal sliding mode control, is used to prove the reachability of the sliding mode surfaces. Simulation examples are given to illustrate the theoretical results.

Corrections to “Finite-Time Attitude Tracking Control for Spacecraft Using Terminal Sliding Model and Chebyshev Neural Network”

Abstract: Errors in the above-named article [ibid., vol. 41, no. 4, pp. 950-963, Aug. 2011] are noted for pages 952, 955, and 956.

Learning Representative Features for Robot Topological Localization

Abstract: This paper proposes a new method for mobile robots to recognize places with the use of a single camera and natural landmarks. In the learning stage, the robot is manually guided along a path. Video sequences are captured with a front-facing camera. To reduce the perceptual alias of visual features, which are easily confused, we propose a modified visual feature descriptor which combines the dominant hue colour information with the local texture. A Location Features Vocabulary Model (LVFM) is established for each individual location using an unsupervised learning algorithm. During the course of travelling, the robot employs each detected interest point to vote for the most likely place. The spatial relationships between the locations, modelled by the Hidden Markov Model (HMM), are exploited to increase the robustness of location recognition in cases of dynamic change or visual similarity. The proposed descriptors are compared with several state-of-the-art descriptors including SIFT, colour SIFT, GLOH and SURF. Experiments show that both the LVFM based on the dominant Hue-SIFT feature and the spatial relationships between the locations contribute considerably to the high recognition rate.

Optimized neural network ensemble by combination of particle swarm optimization and differential evolution

Abstract: The Neural-Network Ensemble (NNE) is a very effective method where the outputs of separately trained neural networks are combined to perform the prediction In this paper, we introduce the improved Neural Network Ensemble (INNE) in which each component forward neural network (FNN) is optimized by particle swarm optimization (PSO) and back-propagation (BP) algorithm At the same time, the ensemble weights are trained by Particle Swarm Optimization and Differential Evolution cooperative algorithm(PSO-DE) We take two obviously different populations to construct our algorithm, in which one population is trained by PSO and the other is trained by DE In addition, we incorporate the fitness value from last iteration into the velocity updating to enhance the global searching ability Our experiments demonstrate that the improved NNE is superior to existing popular NNE

Task-oriented active training based on adaptive impedance control with iLeg — A horizontal exoskeleton for lower limb rehabilitation

Abstract: A horizontal exoskeleton for lower limb rehabilitation called iLeg has been developed by our laboratory which consists of two 3-DOF (degrees of freedom) robotic leg orthoses. This paper proposes a position-based impedance control with the compensation of BP NNs (back propagation neural networks) for the exoskeleton. Based on the control scheme, the task-oriented active training is investigated where impedance parameters are self-adjusted to movement deviation and patient activities with fuzzy logic. An adaptive haptic interface of active compliance is ensured to provide positive feedback to patients when their effort is desired or negative otherwise, which encourages patients to practice the desired movement following the predefined directed path. Besides, the timing freedom is separated from spatial trajectory and determined by patients. Voluntary effort hence becomes a requirement during the exercises, no effort no movement, so that active contribution of patients is highly motivated. Simulation results have verified the feasibility of the control scheme and the training strategy. An active compliant environment is created with adaptive haptic interface for task-oriented patient-driven training of multi-joint coordination.

Correction to "Attitude coordination control for a group of spacecraft without velocity measurements"

Abstract: Manuscript received March 19, 2012; revised April 9, 2012; accepted April 14, 2012. Manuscript received in final form April 20, 2012. Date of publication May 18, 2012; date of current version April 17, 2013. A.-M. Zou and K. D. Kumar are with the Department of Aerospace Engineering, Ryerson University, Toronto, ON M5B 2K3, Canada (e-mail: anmin.zou@ia.ac.cn; kdkumar@ryerson.ca). Z.-G. Hou is with the State Key Laboratory of Intelligent Control and Management of Complex Systems, Institute of Automation, Chinese Academy of Sciences, Beijing 100190, China (e-mail: zengguang.hou@ia.ac.cn). Digital Object Identifier 10.1109/TCST.2012.2196701 Similarly, (45) and (77) should be corrected as

Optimal Proposal Distribution FastSLAM with Suitable Gaussian Weighted Integral Solutions

Abstract: One of the key issues in Gaussian SLAM is to calculate nonlinear transition density of Gaussian prior, ie to calculate Gaussian Weight Integral (GWI) whose integrand is with the form nonlinear function × Gaussian prior density Up to now, some GWI solutions have been applied in SLAM (eg linearization, unscented transform and cubature rule), and different SLAM algorithms were derived based on theirs GWI solutions While, how to select suitable GWI solution for SLAM is still lack of theoretical analysis In this paper, we proposed an optimal proposal FastSLAM algorithm with suitable GWI solutions The main contributions of this work lies that: (1) an unified FastSLAM framework with optimal proposal distribution is summarized; (2) a SLAM dimensionality based GWI solution selection criterion is designed; (3) we propose a new SLAM algorithm The performance of the proposed SLAM is investigated and compared with the FastSLAM20 and UFastSLAM using simulations and our opinion is confirmed by the results