Incorporation of a Regularization Term to Control Negative Correlation in Mixture of Experts

Combining accurate neural networks (NN) in the ensemble with negative error correlation greatly improves the generalization ability. Mixture of experts (ME) is a popular combining method which employs special error function for the simultaneous training of NN experts to produce negatively correlated NN experts. Although ME can produce negatively correlated experts, it does not include a control parameter like negative correlation learning (NCL) method to adjust this parameter explicitly. In this study, an approach is proposed to introduce this advantage of NCL into the training algorithm of ME, i.e., mixture of negatively correlated experts (MNCE). In this proposed method, the capability of a control parameter for NCL is incorporated in the error function of ME, which enables its training algorithm to establish better balance in bias-variance-covariance trade-off and thus improves the generalization ability. The proposed hybrid ensemble method, MNCE, is compared with their constituent methods, ME and NCL, in solving several benchmark problems. The experimental results show that our proposed ensemble method significantly improves the performance over the original ensemble methods.     

A Saturating Extension of an Output Feedback Controller for Internally Damped Euler‐Lagrange Systems

In this research, a novel extension of the passivity‐based output feedback trajectory tracking controller is developed for internally damped Euler‐Lagrange systems with input saturation. Compared with the previous output feedback controllers, this new design of a combined adaptive controller‐observer system will reduce the risk of actuator saturation effectively via generalized saturation functions. Semi‐global uniform ultimate boundedness stability of the tracking errors and state estimation errors is guaranteed by Lyapunov stability analysis. An application of the proposed saturated output feedback controller is the stabilization of a nonholonomic wheeled mobile robot with saturated actuators towards desired trajectories. Simulation results are provided to illustrate the efficiency of the proposed controller in dealing with the actuator saturation.     

Submicron nanoporous polyacrylamide beads with tunable size for verapamil imprinting

Submicron sized polyacrylamide particles were prepared via modified precipitation polymerization method. Experimental design based on Taguchi approach was employed to study the influence of the polymerization composition including monomer (acrylamide), crosslinker (methylenebisacrylamide), initiator (azobisisobutyronitrile), and modifier (polyvinylpyrrolidone, K-30), on the size and morphology of the particles. Varying the polymerization composition, submicron-particles with sizes ranging between 100 and 600 nm were achieved. In all the cases, polydispersity index (PDI) of the particle size was found to be almost 1 indicating uniformity of the particle size. The concentration of crosslinker was found to be the most influential parameter on the particles size and the modifier concentration as an extra tunable parameter was believed to affect the nucleation mechanism and the viscosity of the medium to help controlling the particle size. To validate the optimization, particles with a preset diameter, i.e., 500 nm, were synthesized based on the composition predicted by the mathematical correlation. The polymer with the preset particle size was also imprinted with verapamil and characterized by FTIR, DSC, SEM, physisorption, elemental analysis, swelling, and batch rebinding experiments. The verapamil imprinted polymers bearing nano-cavities exhibited high affinity with imprinting factor 2.17 towards the target molecule. © 2011 Wiley Periodicals, Inc. J Appl Polym Sci, 2012     

Electro-thermo nonlocal nonlinear vibration in an embedded polymeric piezoelectric micro plate reinforced by DWBNNTs using DQM

In the present paper, electro-thermo nonlinear vibration of a piezo-polymeric rectangular micro plate made from polyvinylidene fluoride (PVDF) reinforced by zigzag double walled boron nitride nanotubes (DWBNNTs) is studied. This plate is embedded in an elastic medium which is simulated by Winkler and Pasternak foundation models. Using nonlinear strain-displacement relations and nonlocal elasticity plate theory as well as considering charge equation for coupling between electrical and mechanical fields, the motion equations are derived based on energy method and Hamilton??s principle. The differential quadrature method (DQM) is employed to computation of nonlinear frequency for different mechanical and free-free electrical boundary conditions. The results indicate that smart composite and consequently the generated G4 improved sensor and actuator applications in several process industries, because it increases the nonlinear vibration frequency. Furthermore, it can be also found that the nonlinear frequency increases as the values of the elastic medium constants, the geometrical aspect ratios and DWBNNTs volume fraction increase but it decreases as nonlocal parameter increases.     

Experimental seismic evaluation of old concrete columns reinforced by plain bars

Concrete buildings reinforced by plain (smooth) bars are one of the special types of old reinforced concrete buildings that were generally built before the 1970s. As columns are generally the most important structural members of a framed structure, understanding their realistic seismic behavior is very helpful in estimating structural deformations, forces and energy dissipation capacities. Furthermore, in most of old‐framed building structures, columns play a key role in the final behavior because of strong beam–weak column conditions. This article reports the results of experimental monotonic and cyclic tests on four concrete column specimens reinforced by plain bars and with various types of splices. Through the experimental results, it tried to obtain more clarification on the complicated behavior of such old reinforced members as well as the differences compared with relevant results of columns reinforced by deformed bars. It was realized that slip (fixed‐end rotation) contribution is the major source of deformation in all specimens independent from type of splices. Moreover, general mode of behavior was restrained‐rocking action independent of type of splice detailing. A simple theory for the explanation of hysteresis force–displacement response was proposed. The theory assumes a concrete block rocking element that is restrained with plain bars at both ends. Copyright © 2010 John Wiley & Sons, Ltd.     

Thermal,mechanical, and barrier properties of polyethylene/surlyn/organoclay nanocomposites blown films prepared by different mixing methods

(Low‐density polyethylene) (LDPE)/clay nanocomposites were prepared by melt blending in a twin‐screw extruder by using different mixing methods. Zinc‐neutralized carboxylate ionomer was used as a compatibilizer. Blown films of the nanocomposites were then prepared. The effect of mixing method on the clay dispersion and properties of the nanocomposites was evaluated by wide‐angle X‐ray diffraction analysis, mechanical properties, thermal properties, and barrier properties. The structure and properties of nanocomposites containing different amounts of nanoclay prepared by selected mixing techniques were also investigated. It was found that melt compounding of Surlyn/clay masterbatch with pure LDPE and Surlyn (two‐step‐a method) results in better dispersion and intercalation of the nanofillers than melt mixing of LDPE/Surlyn/clay masterbatch with pure LDPE and surlyn (two‐step‐b method) and direct mixing of LDPE with clay. The films containing ionomer have good barrier properties. A wide‐angle X‐ray diffraction pattern indicates that intercalation of polymer chains into the clay galleries decreases by increasing the clay content. Barrier properties and tensile modulus of the films were improved by increasing the clay content. In addition, tensile strength increased in the machine direction, but it decreased in the transverse direction by increasing the clay content. DSC results showed that increasing the clay content does not show significant change in the melting and crystallization temperatures. The results of thermogravimetric analysis showed that the thermal stability of the nanocomposites decreased by increasing the clay content more than 1 wt%. J. VINYL ADDIT. TECHNOL., 21:60–69, 2015. © 2014 Society of Plastics Engineers     

Investigation of miscibility and phase structure of a novel blend of poly(lactic acid) (PLA)/acrylic rubber (ACM) and its nanocomposite with nanosilica

In this work, a novel polymer blend containing poly(lactic acid) (PLA) as a biocompatible and biodegradable thermoplastic and acrylic rubber (ACM) is prepared and the miscibility and phase structure of the blend and its nanocomposite (PLA/ACM/nanosilica) are investigated through theoretical and experimental methods. To predict the phase diagram of the blend, a compressible regular solution model was employed, in which an upper critical solution temperature was observed. The model predicted that PLA/ACM blends are immiscible over the whole composition range at temperatures below 260 °C. Performing scanning force microscopy on the blend showed phase separated structures for the blends containing different amounts of the PLA and ACM. This was in accordance with the results of dynamic mechanical analysis, which revealed two distinct glass transition temperatures for the studied blends. The effect of nanometer sized silica particle on morphology and rheological properties of these blends was also investigated. Scanning force microscopy results showed much reduction of droplet size in the blends containing 2 wt % nanosilica. This was attributed to the suppression effect of nanosilica on the droplets coalescences. Rheological measurements confirmed the interaction of both components with the silica nanofiller. © 2017 Wiley Periodicals, Inc. J. Appl. Polym. Sci. 2017 , 134, 45499.