Artificial neural networks for modeling in reaction process systems

This work is focused on developing a more efficient computational scheme for estimation of process reaction rates based on NN models. Two scenarios are considered: (1) the kinetics coefficients of the process are completely known and the process states are partly known (measured); (2) the kinetics coefficients and the states of the process are partly known. The contribution of the paper is twofold. From one side we formulate a hybrid (ANN and mechanistic) model that outperforms the traditional reaction rate estimation approaches. From other side, a new procedure for NN supervised training is proposed when target outputs are not available. The two scenarios are successfully tested for two benchmark problems, estimation of the precipitation rate of calcium phosphate and estimation of sugar crystallization growth rate.     

The relation of steady evaporating drops fed by an influx and freely evaporating drops

We discuss a thin film evolution equation for a wetting evaporating liquid on a smooth solid substrate. The model is valid for slowly evaporating small sessile droplets when thermal effects are insignificant, while wettability and capillarity play a major role. The model is first employed to study steady evaporating drops that are fed locally through the substrate. An asymptotic analysis focuses on the precursor film and the transition region towards the bulk drop and a numerical continuation of steady drops determines their fully non-linear profiles. Following this, we study the time evolution of freely evaporating drops without influx for several initial drop shapes. As a result we find that drops initially spread if their initial contact angle is larger than the apparent contact angle of large steady evaporating drops with influx. Otherwise they recede right from the beginning.     

Heuristic dynamic programming using echo state network as online trainable adaptive critic

The present paper proposes an implementation of a relatively new recurrent neural network architecture—the echo state network (ESN)–within the frame of heuristic dynamic programming. The ESN is trained online to estimate the utility function and to adapt the control policy of an embodied agent. With the advantage of an easy training algorithm, the ESN architecture offers a simple way to calculate the derivatives required for adapting the controller. Experimental results are provided to validate the proposed learning approach. Copyright © 2012 John Wiley & Sons, Ltd.     

Mechano mathematical models for mechatronic systems for vibration jaw crushing processes

The aim of this paper is creation, investigation of mono frequency and poly frequency plane and spatial vibrations of two mass vibration systems with a suitable adjustment of the mechanical system. The physical object is a vibrating jaw crusher. The mechanical and mathematical modeling is created on the basis of a two mass dynamic model with six degrees of freedom in plane case—three for each mass and 12 degrees of freedom in spatial model—six for each mass. The influence of the elastic linkages between the two masses and also the energy mutual linkages between the coordinates of the two bodies are investigated.

Natural frequencies and amplitudes of movement are obtained by means of computer experiment in plane and spatial aspect. The obtained results can be used for optimization of the constructive parameters of the vibrating jaw crusher and also for improving harmful resonance effects     

Cover Image,Volume 66, Issue 12

The cover image, by Marianna Triantou et al., is based on the Research Article Mechanical performance of re‐extruded and aged graphene/polypropylene nanocomposites, DOI: 10.1002/pi.5353 .

     

An incremental node embedding technique for error correcting output codes

The error correcting output codes (ECOC) technique is a useful way to extend any binary classifier to the multiclass case. The design of an ECOC matrix usually considers an a priori fixed number of dichotomizers. We argue that the selection and number of dichotomizers must depend on the performance of the ensemble code in relation to the problem domain. In this paper, we present a novel approach that improves the performance of any initial output coding by extending it in a sub-optimal way. The proposed strategy creates the new dichotomizers by minimizing the confusion matrix among classes guided by a validation subset. A weighted methodology is proposed to take into account the different relevance of each dichotomizer. As a result, overfitting is avoided and small codes with good generalization performance are obtained. In the decoding step, we introduce a new strategy that follows the principle that positions coded with the symbol zero should have small influence in the results. We compare our strategy to other well-known ECOC strategies on the UCI database, and the results show it represents a significant improvement.     

Looking Around First: Localized Potential-Based Clustering in Spontaneous Networks

We propose a new budget-based clustering algorithm for self-organizing networks. The basic idea behind our solution is that nodes first sense the environment using inherent Hello packets before starting forming clusters. In contrast with previous solutions that blindly distribute the budget to nearby nodes, our potential-based clustering algorithm applies a proportional budget distribution based on the connectivity degree of the nodes. This approach matches the principle that nodes in real networks are not uniformly distributed. Our simulation results show that the proposed approach outperforms previous ones.     

Distinct Bimodal Roles of Aromatic Molecules in Controlling Gold Nanorod Growth for Biosensing

New aromatic molecule–seed particle interactions are examined and exploited to control and guide seed‐mediated gold nanorod (Au NR) growth. This new approach enables better understanding of how small molecules impact the synthesis of metallic nanostructures, catalyzing their use in various biomedical applications, such as plasmonic biosensing. Experimental studies and theoretical molecular simulations using a library of aromatic molecules, making use of the chemical versatility of the molecules with varied spatial arrangements of electron‐donating/withdrawing groups, charge, and Au‐binding propensity, are performed. Au NR growth is regulated by two principal mechanisms, producing either a red or blue shift in the longitudinal localized surface plasmon resonance (LLSPR) peaks. Aromatic molecules with high redox potentials produce an increase in NR aspect ratio and red shift of LLSPR peaks. In contrast, molecules that strongly bind gold surfaces result in blue shifts, demonstrating a strong correlation between their binding energy and blue shifts produced. Through enzymatic conversion of selected molecules, 4‐aminophenylphosphate to 4‐aminophenol, opposing growth mechanisms at opposite extremes of target concentration are obtained, and a chemical pathway for performing plasmonic enzyme‐linked immunosorbent assays is established. This unlocks new strategies for tailoring substrate design and enzymatic mechanisms for controlling plasmonic response to target molecules in biosensing applications.