Discrete cuckoo search algorithms for two-sided robotic assembly line balancing problem

Neural Computing and Applications - Robotics are extensively utilized in modern industry to replace human labor and achieve high automation and flexibility. In order to produce large-size products,...     

Control of a Mobile Robot Subject to Wheel Slip

Wheel slip is inevitable when a Wheeled Mobile Robot (WMR) is moving at a high speed or on a slippery surface. In particular, when neither lateral nor longitudinal slips can be ignored in the dynamic model, a WMR becomes an under-actuated nonlinear dynamic system. To study the maneuverability of a WMR in such a realistic environment, we model the overall WMR dynamics subject to wheel slip and propose control algorithms in regulation control and turning control tasks for the WMR. In regulation control, a time-invariant discontinuous feedback law is developed to asymptotically stabilize the system to the desired configuration with exponential convergence rate. In turning control, a sliding mode-based extremum seeking control technique is applied to achieve stable and sharp turning. Simulation results are presented to validate the theoretical results.     

Models for the behavior of boron carbide in extreme dynamic environments

We describe models for the behavior of hot-pressed boron carbide that is subjected to extreme dynamic environments such as ballistic impact. We first identify the deformation and failure mechanisms that are observed in boron carbide under such conditions, and then review physics-based models for each of these mechanisms and the integration of these models into a single physics-based continuum model for the material. Atomistic modeling relates the composition and stoichiometry to the amorphization threshold, while mesoscale modeling relates the processing-induced defect distribution to the fracture threshold. The models demonstrate that the relative importance of amorphization and fracture are strongly dependent on the geometry and impact conditions, with the volume fraction of amorphized material being unlikely to be significant until very high velocities (~3 km/s) are reached for geometries such as ball impact on plates. These connections to the physics thus provide guidelines for the design of improved boron carbide materials for impact applications.     

Effects of Lewis number on turbulent scalar transport and its modelling in turbulent premixed flames

The behaviour of the turbulent scalar flux in premixed flames has been studied using Direct Numerical Simulation (DNS) with emphasis on the effects of Lewis number in the context of Reynolds-averaged closure modelling. A database was obtained from DNS of three-dimensional freely propagating statistically planar turbulent premixed flames with simplified chemistry and a range of global Lewis numbers from 0.34 to 1.2. Under the same initial conditions of turbulence, flames with low Lewis numbers are found to exhibit counter-gradient transport, whereas flames with higher Lewis numbers tend to exhibit gradient transport. The Reynolds-averaged transport equation for the turbulent scalar flux is analysed in detail and the performance of existing models for the unclosed terms is assessed with respect to corresponding quantities extracted from DNS data. Based on this assessment, existing models which are able to address the effects of non-unity Lewis number on turbulent scalar flux transport are identified, and new or modified models are suggested wherever necessary. In this way, a complete set of closure models for the scalar flux transport equation is prescribed for use in Reynolds-Averaged Navier-Stokes simulations.     

The effects of turbulence on molten pool transport during melting and solidification processes in continuous conduction mode laser welding of copper–nickel dissimilar couple

The melting and solidification stages of a continuous copper–nickel dissimilar metal conduction mode laser welding have been simulated numerically in this study. The heat, mass and momentum transports in molten metal pool have been analysed using both laminar and turbulent flow models separately for the same process parameters. The phase change aspects related to solidification and melting are accounted for by a modified enthalpy–porosity technique while the turbulent transport is modelled by a high Reynolds number k–ε model. It has been observed that temperature fields obtained from both laminar and turbulent transport simulations are qualitatively similar to each other. The molecular thermal diffusivity of the molten metal mixture is found to be in the same order of magnitude as eddy thermal diffusivity, as a result of which the thermal field gets marginally affected by fluid turbulence. By contrast, eddy viscosity remains much greater than molecular viscosity, which leads to greater amount of momentum diffusion in the case of a turbulent molten metal pool, in comparison to that obtained from the corresponding laminar simulation. This is reflected in the reduction in maximum velocity magnitude in the turbulent simulation in comparison to the maximum velocity obtained from laminar simulation. In the case of species transport, the turbulent mass diffusivity is found to be about 10⁷–10⁸ times greater than molecular mass diffusivity. As a result, the species field in turbulent simulation shows characteristics of better mixing between two dissimilar molten metals than the species field obtained using the laminar transport model. The species distribution obtained from turbulent transport is shown to be in better agreement with experimental data reported in literature than the corresponding mass fraction distribution obtained from laminar simulation. It is also found that species distribution in the molten pool is principally determined by advective and diffusive transport during the melting stage and species transport by advection and eddy diffusion in turbulent pool increasingly weakens with decreasing temperature during the cooling following the laser melting stage.     

Intelligent stuttering speech recognition: A succinct review

Stuttering speech recognition is a well-studied concept in speech signal processing. Classification of speech disorder is the main focus of this study. Classification of stuttered speech is becoming more important with the enhancement of machine learning and deep learning. In this study, some of the recent and most influencing stuttering speech recognition methods are reviewed with a discussion on different categories of stuttering. The stuttering speech recognition process is divided mainly into four segments-input speech pre-emphasis, segmentation, feature extraction, and stutter classification. All these segments are briefly elaborated and related researches are discussed. It is observed that different traditional machine learning and deep learning classification approaches are employed to recognize stuttered speech in last few decades. A comprehensive analysis is presented on different feature extraction and classification method with their efficiency.

     

Robust nonlinear analytic redundancy for fault detection and isolation in mobile robot

A robust nonlinear analytical redundancy (RNLAR) technique is presented to detect and isolate actuator and sensor faults in a mobile robot. Both model-plant-mismatch (MPM) and process disturbance are considered during fault detection. The RNLAR is used to design primary residual vectors (PRV), which are highly sensitive to the faults and less sensitive to MPM and process disturbance, for sensor and actuator fault detection. The PRVs are then transformed into a set of structured residual vectors (SRV) for fault isolation. Experimental results on a Pioneer 3-DX mobile robot are presented to justify the effectiveness of the RNLAR scheme.     

Transport equations for reaction rate in laminar and turbulent premixed flames characterized by non-unity Lewis number

Transport equations for (i) the rate $W$ of product creation and (ii) its Favre-averaged value $\stackrel{?}{W}$ are derived from the first principles by assuming that $W$ depends solely on the temperature and mass fraction of a deficient reactant in a premixed turbulent flame characterized by the Lewis number $Le$ different from unity. The right hand side of the transport equation for $\stackrel{?}{W}$ involves seven unclosed terms, with some of them having opposite signs and approximately equal large magnitudes when compared to the left-hand-side terms. Accordingly, separately closing each term does not seem to be a promising approach, but a joint closure relation for the sum $\overline{T_{\text{Σ}}}$ of the seven terms is sought. For this purpose, theoretical and numerical investigations of variously stretched laminar premixed flames characterized by $Le<1$ are performed and the linear relation between $T_{\text{Σ}}$ integrated along the normal to a laminar flame and a product of (i) the consumption velocity $u_c$ and (ii) the stretch rate ${\dot{s}}_w$ evaluated in the flame reaction zone is obtained. Based on this finding and simple physical reasoning, a joint closure relation of $\overline{T_{\text{Σ}}}\propto \overline{\rho W\dot{s}}$ is hypothesized, where $\rho$ is the density and $\dot{s}$ is the stretch rate. The joint closure relation is tested against 3D DNS data obtained from three statistically 1D, planar, adiabatic, premixed turbulent flames in the case of a single-step chemistry and $Le=0.34$, 0.6, or 0.8. In all three cases, the agreement between $\overline{T_{\text{Σ}}}$ and $\overline{\rho W\dot{s}}$ extracted from the DNS is good with exception of large ($\overline{c}>0.4$) values of the mean combustion progress variable $\overline{c}$ in the case of $Le=0.34$. The developed linear relation between $\overline{T_{\text{Σ}}}$ and $\overline{\rho W\dot{s}}$ helps to understand why the leading edge of a premixed turbulent flame brush can control its speed.     

Short‐term extreme response analysis of a jacket supporting an offshore wind turbine

Wind turbines must be designed in such a way that they can survive in extreme environmental conditions. Therefore, it is important to accurately estimate the extreme design loads. This paper deals with a recently proposed method for obtaining short‐term extreme values for the dynamic responses of offshore fixed wind turbines. The 5 MW NREL wind turbine is mounted on a jacket structure (92 m high) at a water depth of 70 m at a northern offshore site in the North Sea. The hub height is 67 m above tower base or top of the jacket, i.e. 89 m above mean water level. The turbine response is numerically obtained by using the aerodynamic software HAWC2 and the hydrodynamic software USFOS . Two critical responses are discussed, the base shear force and the bending moment at the bottom of the jacket. The extreme structural responses are considered for wave‐induced and wind‐induced loads for a 100 year return‐period harsh metocean condition with a 14.0 m significant wave height, a 16 s peak spectral period, a 50 m s ^{? 1} (10 min average) wind speed (at the hub) and a turbulence intensity of 0.1 for a parked wind turbine. After performing the 10 min nonlinear dynamic simulations, a recently proposed extrapolation method is used for obtaining the extreme values of those responses over a period of 3 h. The sensitivity of the extremes to sample size is also studied. The extreme value statistics are estimated from the empirical mean upcrossing rates. This method together with other frequently used methods (i.e. the Weibull tail method and the global maxima method) is compared with the 3 h extreme values obtained directly from the time‐domain simulations. Copyright © 2012 John Wiley & Sons, Ltd.