In order to realize the non-destructive intelligent identification of weld surface defects, an recognition method based on deep learning is proposed, which mainly formed by convolutional neural network (CNN) and forest random. First, high-level features are automatically learned through CNN. Random trained with extracted predict classification results. Secondly, defects images collected preprocessed image enhancement threshold segmentation. A database established using pre-processed images....

Geometric features of the phase diagram a cantilever beam with breathing crack are investigated. Influences properties on geometry studied and method for identification based is proposed. The pixel used to estimate different areas. Multiple indexes defined under conditions primary super-harmonic resonance. effects location depth each index analyzed. contour lines multiple parameters. Numerical calculations indicate that diagrams resonance more sensitive parameters proposed both feasible effective.

With the development of exoskeleton rehabilitation robotic system, biomechanics influence device should be considered in training. This paper describes a novel upper limb (ULERD) based on biomechanics, which can used for hemiparalysis patients. system is aimed at helping patients recover motor function and are suitable variety portable wearable, consists device, haptic (PHANTOM Premium), computer. By using finite element software ANSYS, patient's training explored. Based it, optimization...

For the extreme non-linear and strong coupling characteristics of mobile manipulators, based on theory sliding mode variable structure control, this paper proposes a control approach manipulators. The entire system is composed two parts, including trajectory tracking platform non-singular terminal manipulator. simulation results show effectiveness presented approach.

In this paper, a novel over-constrained three degree-of-freedom (DOF) spatial parallel manipulator (SPM) is proposed. The architecture of the SPM comprised moving platform attached to base through two revolute- prismatic-universal jointed serial linkages and one spherical-prismatic-revolute linkage. prismatic joints are considered be actively actuated. Kinematics dynamics studied systematically. Firstly, both inverse forward displacements derived by analytic formulae. Then, velocities...

SUMMARY Double inverted pendulum model, stationary or on a cart, is computationally the simplest out of range reasonable models used for anthropomorphic robots motion synthesis. However, it still not sufficient describing more complex situations. The novel concept variable double (VDIP) static postures and VDIP cart (VDIPC) dynamic cases proposed. It provides simplified but sufficiently accurate tool planning human-like robot postures. Its parameters enable description both human dynamics....

This paper discusses the issues of dynamic parameter identification for reconfigurable robot and presents an approach based on Adaline neural network. The process is mainly divided into three steps. first to establish model compensate joint friction with Coulomb viscous model. Moreover, finite Fourier series designed by excitation trajectory genetic algorithm used optimize trajectory. second build network train weight motion torques information robot. third simulate three-degree-of-freedom...

<para xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink"> Some physics researchers retrieve EAST experiment data using the Transmission Control Protocol (TCP) in a wireless local area network (WLAN). TCP is most commonly used transport control protocol. It assumes that every packet loss caused by congestion and invokes avoidance. TCP's blind results degraded performance lossy networks. In network, mobile hosts have first-hand knowledge of links;...

Two types of 3-CRU translational parallel manipulators (TPM) are compared and investigated in this paper. The two TPMs have identical kinematics, but some differences exist terms the chain arrangement, one is fully symmetrical called TPM, other asymmetrical TPM. This paper focuses on discussing between workspace performance. study provides insights into with kinematics arranged differently.

A unit module of the mobile modular reconfigurable robot (M2rBot) with 12 connection ways is first introduced. For expressing spatial relations and traversing paths M2rBot, a novel method automatic generation matrix presented based on our proposed double-module space pose transformation. The motion equations are also automatically generated to solve reconstruction problems mechanical configuration, kinematics, dynamics. This enables analyze kinematics dynamics for both serial multi-branch...

The paper proposes a solution to an inverse kinematics problem based on dual quaternions algebra. method, relying screw theory, requires less calculation effort compared with commonly used approaches. obtained kinematic description is very concise, and the singularity avoided. formalism applied decomposition description. As example, of multi-DOF serial manipulator considered. Direct problems are solved using division into sub-problems. Each new sub-problem proposed concerned rotation about...

An innovative modular self-reconfigurable robot is presented, namely 360bot, which has a continuous rotational degree of freedom.In addition to the common features conventional robots, each module can move independently and reliably due its special simple structure, thereafter it imposes less restriction on surrounding environment.The mechanism designed with multiple connections, including four connected faces five connection modes, enables unit possessing multi-directional degrees freedom,...

Based on kinematic and kinetic analyses of the 3-degree freedom differentially driven wheels mobile robot, we proposed a nonlinear mathematical model with uncertain disturbance. This was transformed into linear control systems through an approximate linearization algorithm, which is based balanced flow pattern. After that, designed partial feedback H <sub xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">∞</sub> robust controller LMI for robots. The...

Fuzzing is a widely-used software vulnerability discovery technology, many of which are optimized using coverage-feedback.Recently, some techniques propose to train deep learning (DL) models predict the branch coverage an arbitrary input owing its always-available gradients etc. as guide.Those have proved their success in improving and discovering bugs under different experimental settings.However, DL models, usually magic black-box, notoriously lack explanation.Moreover, performance can be...

This paper proposes a sliding mode control approach for wheeled mobile manipulator. The manipulator composed of four-wheeled platform and three-degree freedom onboard dynamic models are established the methods employed. entire system is decomposed into two subsystems, including non-singular terminal simulation results shown at last demonstrate effectiveness presented algorithm.

A kinematic modeling method based on the dual quaternion is proposed. The requires less calculation time comparing to traditional approaches, e.g. approach using Denavit-Hartenberg formalism. forward problem for 6-DOF manipulator derived quaternions. When solving inverse problem, decomposition Paden-Kahan sub-problem applied. new formulated as a "rotation about two subsequent axes given amount" In general, with this method, of multi-DOF decomposed into several sub-problems. 6-degree freedom...

To detect a transverse crack caused by fatigue or creep, most of the research has thus far paid attention only to resonances steady-state oscillations created and proposed diagnosis systems utilizing these vibration phenomena.However, from practical view point, have following flaws: (1) probability that resonance occurs due in rated rotational speed range is lower position; (2) It very dangerous observe characteristics ranges.In order solve problems, this paper uses detection method...

Two-dimensional principal component analysis algorithm (2DPCA) can be performed in the batch mode and not meet real-time requirements of video stream. To overcome these limitations, incremental learning candid covariance-free PCA (CCIPCA) is innovated to existing 2DPCA, called 2DPCA (I2DPCA) firstly presented incrementally compute components a sequence samples directly on 2D image matrices without estimating covariance matrices. Therefore, I2DPCA improve feature extraction speed reduce...

This paper aims at solving the problems that manipulator will be constrained by flexible load, and caused constraint force is difficult to control. Based on deflection screw theory, dynamic model of system established using Lagrange equation. Firstly, according interaction end effector, load environmental constraints manipulator, equation motion entire built. Secondly, set up, including manipulator. Finally, Under certain known constraints, modeling a performed. The effectiveness proposed...

To guarantee the quality of software, specifying security requirements (SRs) is essential for developing systems, especially security-critical software systems. However, using threat to determine detailed SR quite difficult according Common Criteria (CC), which too confusing and technical non-security specialists. In this paper, we propose a Co-occurrence Recommend Model (CoRM) automatically recommend SRs. model, threats product are extracted from target documents in related tagged. order...