Prediction dielectric constant of different ternary liquid mixtures at various temperatures and compositions using artificial neural networks

Artificial neural networks (ANNs) were successfully developed for the modeling and prediction dielectric constant of different ternary liquid mixtures at various temperatures (?10°C?≤?t?≤?80°C) and over the complete composition range (0?≤?x ₁,?x ₂,?x ₃?≤?1). A three-layered feed forward ANN with architecture 7-16-1 was generated using seven parameters as inputs and its output is dielectric constant of media. It was found that properly selected and trained neural network could fairly represent the dependence of dielectric constant of different ternary liquid mixtures on temperature and composition. For the evaluation of the predictive power of the generated ANN, an optimized network was applied for predicting the dielectric constant in the prediction set, which were not used in the modeling procedure. Squared correlation coefficient (R ²) and root mean square error for prediction set are 0.9997 and 0.2060, respectively. The mean percent deviation (MPD) for the property in the prediction set is 0.8892%. The results show nonlinear dependence of dielectric constant of ternary mixed solvent systems on temperature and composition is significant.     

Optimization and empirical modeling of HG-ICP-AES analytical technique through artificial neural networks.

An artificial neural network technique has been applied to the optimization of a hydride generation-inductively coupled plasma-atomic emission spectrometry (HG-ICP-AES) coupling for the determination of Ge at trace levels. The back propagation of errors net architecture was used. Experimental parameters and their relationship have been studied, obtaining a surface response of the system. The results and optimization aspects achieved with the neural network approach have been compared to the "one variable at time" and SIMPLEX methods.     

Detection of liner surface defects in solid rocket motors using multilayer perceptron neural networks

Debonding problems along the propellant/liner/insulation interface are a critical point to the integrity and one of the major causes of structural failures of solid rocket motors. Current solutions are typically restricted to methods for assessing the integrity of the rocket motors structure and visually inspecting their components. In this context, this paper presents an improved algorithm to detect liner surface defects that may compromise the bonding between the solid propellant and the insulation. The use of Local Binary Patterns (LBP) provides a structural and statistical approach to texture analysis of liner sample images. Along with color information extraction, these two methods allow the representation of image pixels by feature vectors that are further processed by a Multilayer Perceptron (MLP) neural network classifier. The MLP neural network analyzes liner sample images and classifies each pixel into one of three classes: non-defect, foreign object, and defect. Several tests were executed varying different parameters to find the optimal MLP configuration, and as a result, the best classification accuracy of 99.08%, 90.66%, and 99.48% was achieved for the corresponding classes. Moreover, the defect size estimate showed that the MLP classifier correctly identified defects less than 1 mm long, with a relatively small number of training examples. Positive results indicate that the algorithm can identify liner surface defects with a performance similar to human inspectors and has the potential to assist or even automate the liner inspection process of solid rocket motors.     

Using artificial neural networks for prediction of organic acid partition coefficients

Aqueous/organic phase partition coefficients of organic acids were predicted using an artificial neural network (ANN) algorithm taking benzoic acid derivatives as examples. The partition coefficients were determined by extraction of the acids from aqueous salt solutions with hydrophilic solvents (BuⁿOH, BuⁱOH, and Bu^tOH). Using the ANN approach makes it possible to obtain quantitative information on the values of the title parameters. Published in Russian in Izvestiya Akademii Nauk. Seriya Khimicheskaya, No. 2, pp. 207—212, February, 2006.     

Optimization of extraction of linarin from Flos chrysanthemi indici by response surface methodology and artificial neural network

The extraction of linarin from Flos chrysanthemi indici by ethanol was investigated. Two modeling techniques, response surface methodology and artificial neural network, were adopted to optimize the process parameters, such as, ethanol concentration, extraction period, extraction frequency, and solvent to material ratio. We showed that both methods provided good predictions, but artificial neural network provided a better and more accurate result. The optimum process parameters include, ethanol concentration of 74%, extraction period of 2 h, extraction three times, solvent to material ratio of 12 mL/g. The experiment yield of linarin was 90.5% that deviated less than 1.6% from that obtained by predicted result.     

Simultaneous kinetic determination of sulfite and sulfide using artificial neural networks

Artificial neural networks (ANNs) are proposed for the determination of sulfite and sulfide simultaneously. The method is based on the reaction between Brilliant Green (BG) as a colored reagent and sulfite and/or sulfide in buffered solution (pH 7.0) and monitoring the changes of absorbance at maximum wavelength of 628 nm. Experimental conditions such as pH, reagents concentrations, and temperature were optimized and training the network was performed using principal components (PCs) of the original data. The network architecture (number of input, hidden and output nodes), and some parameters such as learning rate (η) and momentum (α) were also optimized for getting satisfactory results with minimum errors. The measuring range was 0.05-3.6 μg ml⁻¹ for both analytes. The proposed method has been successfully applied to the quantification of the sulfite and sulfide in different water samples.     

Solvent effects on kinetics of an aromatic nucleophilic substitution reaction in mixtures of an ionic liquid with molecular solvents and prediction using artificial neural networks

Kinetics of the reaction between 1‐chloro‐2,4‐dinitrobenzene and aniline was studied in mixtures of 1‐ethyl‐3‐methylimidazolium ethylsulfate ([EMIM][EtSO₄]) with methanol, chloroform, and dimethylsulfoxide at 25°C. Single‐parameter correlations of log k_A versus normalized polarity parameter (E^N_T), hydrogen‐bond acceptor basicity (β), hydrogen‐bond donor acidity (α), and dipolarity/polarizability (π*) of media do not give acceptable results. Multiparameter linear regression (MLR) of log k_A versus the solvatochromic parameters demonstrates that the reaction rate constant increases with E^N_T, π*, and β and decreases with α parameter. To predict accurately solvent effects on the rate constant, optimized artificial neural network with three inputs (including α, π*, and β parameters) was applied for prediction of the log k_A values in the prediction set. It was found that properly selected and trained neural network could fairly represent the dependence of the reaction rate constant on solvatochromic parameters. Mean percent deviation of 5.023 for the prediction set by the MLR model should be compared with the value of 0.343 by the artificial neural network model. These improvements are due to the fact that the reaction rate constant shows nonlinear correlations with the solvatochromic parameters. © 2008 Wiley Periodicals, Inc. Int J Chem Kinet 41: 153–159, 2009     

Stress-optical coefficient of cis-1,4-polybutadiene and cis-1,4-polyisoprene networks. Measurements on cis-1,4-polybutadiene networks and theoretical interpretation

Stress, strain, and birefringence measurements have been carried out on swollen and unswollen networks of ′cis-1,4-polybutadiene polymers. Neither stress-strain nor birefringence-strain relations of unswollen specimens obey the Gaussian network theory, but both can be fitted by the Mooney-Rivlin equation. On the contrary, data on specimens swollen in tetralin, decalin, benzene, and carbon tetrachloride strictly obey the Gaussian network theory. Existing methods for evaluating the temperature coefficient of the unperturbed dimensions, d In 〈r²〉/dT, from the stress-temperature relation are applied to the present data and discussed in some detail. It is concluded that reliable values of d In 〈r²〉/dT are not obtainable from data on unswollen samples because of the pronounced non-Gaussian effect. The value 7.5 ų for the optical anisotropy ų (an alternative to the stress-optical coefficient) for unswollen specimens is markedly larger than values (5.8 ų on the average) for swollen specimens. This is interpreted as due to the shortrange orientational order among polymer segments. The quantities 〈r²〉, ΔΓ, and their temperature coefficients are calculated for both cis-1,4-polybutadiene and cis-1,4-polyisoprene chains, on the basis of the rotational isomeric state approximation for bond rotations. Values of ΔΓ for cis-1,4-polybutadiene calculated using Clément and Bothorel's set of anisotropic bond polarizabilities are in good agreement with observed values for swollen specimens. Those for cis-1,4-polyisoprene obtained using the same set of anisotropic bond polarizabilities are somewhat smaller than observed values for unswollen specimens. This departure is in the direction expected from the behavior of ΔΓ upon swelling (i.e., a decrease in ΔΓ upon swelling).     

Entanglement networks of 1,2-polybutadiene crosslinked in states of strain. VII. Stress-birefringence relations

Crosslinks are introduced by γ irradiation into 1,2-polybutadiene while strained in uniaxial extension near T_g with stretch ratio λ₀, thereby trapping a proportion of the entanglements originally present. The stress at any subsequent strain λ is accurately given by the sum σ_N + σ_x, where σ_N is the stress contributed by a trapped entanglement network with λ = 1 as reference and a Mooney–Rivlin stress-strain relation, and σ_x is that contributed by a crosslink network with λ = λ₀ as reference and neo-Hookean stress-strain relation. The birefringence is accurately given as δn = ?_Nσ_N + ?_xσ_x, where the ?'s are the respective stress-optical coefficients. From measurements at λ = λ₀ where σ_x = 0, ?_N can be determined separately. For polymer with 88% 1,2 microstructure, ?_N and ?_x are nearly equal and independent of irradiation dose, though strongly dependent on temperature. For polymer with (95–96)% 1,2, ?_N and ?_x are different (even opposite in sign) and dependent on dose. This behavior is associated with a side reaction of cyclization by the γ irradiation, which is inhibited by the 1,4 moiety in the polymer with lesser 1,2 content. It is responsible for residual birefringence in the state of ease (λ = λ_s) where σ_N = –σ_x and the stress is zero.     

Dynamics of supramolecular associative polymer networks at the interplay of chain entanglement,transient chain association,and chain‐sticker clustering

The dynamic mechanical properties of supramolecular associative polymer networks depend on the average number of entanglements along the network‐forming chains, N_e, and on their content of associative groups, f . In addition, there may be further influence by aggregation of the associative groups into clusters, which, in turn, is influenced by the chemical structure of these groups, and again by N_e and f of the polymer. Therefore, the effects of these parameters are interdependent. To conceptually understand this interdependency, we study model networks in which (a) N_e, (b) f , and (c) the chemical structure of the associative groups are varied systematically. Each network is probed by rheology. The clustering of the associative groups is assessed by analyzing the rheological data at the end range of frequency covered and by comparison of the number of supramolecular network junctions with the maximum possible number of binary transient bonds. We find that if the total number of the network junctions, which can be formed either by interchain entanglement or by interchain transient associations, is greater than a threshold of 13, then the likelihood of cluster formation is high and the dynamics of supramolecular associative polymer networks is mainly controlled by this phenomenon. © 2019 Wiley Periodicals, Inc. J. Polym. Sci., Part B: Polym. Phys. 2019 , 57, 1209–1223     

Dynamic - mechanical properties of poly(oxypropylene)di-amine-diepoxide and poly(oxypropylene)triamine-diepoxide networks and their relationship to the structure of elastically active network chains

The effect of the initial mole ratio of reactive components on the shape and position of dynamic mechanical functions in the main transition and rubbery region was investigated for two series of networks made from poly(oxypropylene)diamine (D-400)-diglycidyl ether of Bisphenol A (DGEBA) and poly(oxypropylene)-triamine (T-403)-DGEBA. The networks were prepared with an excess of amine groups up to the highest conversion of epoxy groups; the ratio ^rH = 2 [ NH₂ ]₀/ [E]₀ ranged from unity to 2,1 for networks from D-400 and from unity to 3,5 for networks from T-403. By using the theory of branching processes, structural parameters of these networks were calculated, in particular, the molecular weights of elastically active network chains (EANC's) including dangling chains, of backbone EANC's and of dangling chains. A comparison between theory and experiment led to the following conclusions: (a) the mechanical behaviour in the rubberlike region can be described either by using an affine deformation model (front factor A = 1), or by means of a phantom model (A = (f_e-2)/f_e, f_e being functionality of the crosslink) with the contribution of permanent interchain interactions; (b) the temperature and frequency position of viscoelastic functions in the main transition region is conclusively affected by the concentration of EANC's; (c) the shape of visco-elastic functions, especially of retardation spectra in the main transition and rubbery region, depends on the detailed structure of EANC's, but it cannot be decided from the result which structural parameter has the strongest effect on the shape of the functions.     

Application of neural networks with novel independent component analysis methodologies for the simultaneous determination of cadmium,copper, and lead using an ISE array

The paper introduces a novel chemometric strategy based on independent component analysis (ICA) coupled with a back‐propagation neural network. In this approach, one of the most popular ICA methods, the fast fixed‐point algorithm for ICA (fastICA), was implemented by the genetic algorithm (geneticICA) to avoid the local maxima problem commonly observed with fastICA. As a case study, an ion‐selective electrode (ISE) array, consisting of three working electrodes and one reference electrode, was used for the simultaneous determination of three heavy metals (cadmium, copper, and lead) in aqueous solutions, which are normally prone to severe interferences. The robustness and appropriateness of the approach were assessed using the average mean of relative error (MRE) of triplicated external validation. After configuration and optimization, the average MRE for Cu was <5%. For the determination of Cd and Pb, whose ISEs normally cannot tolerate Cu ions even at the microgram per liter levels, the MREs were 8%. This article demonstrated that this approach can be applied to the detection of heavy metal contamination in industrial wastewater with prediction accuracies comparable with other popular quantitative chemometric neural network methods. Copyright © 2014 John Wiley & Sons, Ltd.     

Epoxide networks as model networks

Summary The equilibrium mechanical behaviour of weak diepoxide-monoepoxide-diamine networks, prepared with an excess of diamine and measured in the rubbery state, was compared with theoretical predictions obtained by using the theory of branching processes. The experimental equilibrium moduli fit well the shape of theoretical curves over a broad range of crosslinking density regardless of whether the contribution by trapped entanglements is considered or not. The data fit equally well the theoretical dependence for the front factor A = 1 without entanglement contribution and forA = (f _e - 2)/f _e , (f _e is the average effective functionality of a junction) with an entanglement contribution based on the contact probability between any two units within elastically active network chains (Langley).Dedicated to Prof. Dr.G. Rehage on the occasion of his 60th birthday.With 5 figures     

Hybridized particle swarm algorithm for adaptive structure training of multilayer feed-forward neural network: QSAR studies of bioactivity of organic compounds

The multilayer feed-forward ANN is an important modeling technique used in QSAR studying. The training of ANN is usually carried out only to optimize the weights of the neural network and without paying attention to the network topology. Some other strategies used to train ANN are, first, to discover an optimum structure of the network, and then to find weights for an already defined structure. These methods tend to converge to local optima, and may also lead to overfitting. In this article, a hybridized particle swarm optimization (PSO) approach was applied to the neural network structure training (HPSONN). The continuous version of PSO was used for the weight training of ANN, and the modified discrete PSO was applied to find appropriate the network architecture. The network structure and connectivity are trained simultaneously. The two versions of PSO can jointly search the global optimal ANN architecture and weights. A new objective function is formulated to determine the appropriate network architecture and optimum value of the weights. The proposed HPSONN algorithm was used to predict carcinogenic potency of aromatic amines and biological activity of a series of distamycin and distamycin-like derivatives. The results were compared to those obtained by PSO and GA training in which the network architecture was kept fixed. The comparison demonstrated that the HPSONN is a useful tool for training ANN, which converges quickly towards the optimal position, and can avoid overfitting in some extent.     

Multi-objective simultaneous prediction of waterborne coating properties

Multi-objective simultaneous prediction of waterborne coating properties was studied by the neural network combined with programming. The conditions of network with one input layer, three hidden layers and one output layer were confirmed. The monomers mass of BA, MMA, St and pigments mass of TiO₂ and CaCO₃ were used as input data. Four properties, which were hardness, adhesion, impact resistance and reflectivity, were used as network output. After discussing the hidden layer neurons, learn rate and the number of hidden layers, the best net parameters were confirmed. The results of experiment show that multi-hidden layers was advantageous to improve the accuracy of multi-objective simultaneous prediction. 36 kinds of coating formulations were used as the training subset and 9 acrylate waterborne coatings were used as testing subset in order to predict the performance. The forecast error of hardness was 8.02% and reflectivity was 0.16%. Both forecast accuracy of adhesion and impact resistance were 100%.