Performance comparison of feedforward neural network training algorithms in modeling for synthesis of polycaprolactone via biopolymerization

This paper reports the biopolymerization of ε-caprolactone, using lipase Novozyme 435 catalyst at varied impeller speeds and reactor temperatures. A multilayer feedforward neural network (FFNN) model with 11 different training algorithms is developed for the multivariable nonlinear biopolymerization of polycaprolactone (PCL). In previous works, biopolymerization carried out in scaled-up bioreactors is modeled through FFNN. No review discussed the role of different training algorithms in artificial neural network on the estimation of biopolymerization performance. This paper compares mean absolute error, mean square error, and mean absolute percentage error (MAPE) in the PCL biopolymerization process for 11 different training algorithms that belong to six classes, namely (1) additive momentum, (2) self-adaptive learning rate, (3) resilient backpropagation, (4) conjugate gradient backpropagation, (5) quasi-Newton, and (6) Bayesian regulation propagation. This paper aims to identify the most effective training method for biopolymerization. Results show that the quasi-Newton-based and Levenberg–Marquardt algorithms have the best performance with MAPE values of 4.512, 5.31, and 3.21% for the number of average molecular weight, weight average molecular weight, and polydispersity index, respectively.     

Automatic Sentimental Analysis by Firefly with Levy and Multilayer Perceptron

The field of sentiment analysis (SA) has grown in tandem with the aid of social networking platforms to exchange opinions and ideas. Many people share their views and ideas around the world through social media like Facebook and Twitter. The goal of opinion mining, commonly referred to as sentiment analysis, is to categorise and forecast a target’s opinion. Depending on if they provide a positive or negative perspective on a given topic, text documents or sentences can be classified. When compared to sentiment analysis, text categorization may appear to be a simple process, but number of challenges have prompted numerous studies in this area. A feature selection-based classification algorithm in conjunction with the firefly with levy and multilayer perceptron (MLP) techniques has been proposed as a way to automate sentiment analysis (SA). In this study, online product reviews can be enhanced by integrating classification and feature election. The firefly (FF) algorithm was used to extract features from online product reviews, and a multi-layer perceptron was used to classify sentiment (MLP). The experiment employs two datasets, and the results are assessed using a variety of criteria. On account of these tests, it is possible to conclude that the FFL-MLP algorithm has the better classification performance for Canon (98% accuracy) and iPod (99% accuracy).     

Preliminary studies on molybdenum-doped indium oxide thin films deposited by radio-frequency magnetron sputtering at room temperature

Thin films of molybdenum-doped indium oxide (IMO) were prepared by a 3-source, cylindrical radio-frequency magnetron sputtering at room temperature. The films were post-annealed and were characterized by their structural (X-ray diffraction) and optical (UV-VIS-NIR spectrophotometer) properties. The films were studied as a function of oxygen volume percentage (O₂ vol.%) ranging from 3.5 to 17.5. The structural studies revealed that the as-deposited amorphous films become crystalline on annealing. In most cases, the (222) reflection emerged as high intensive peak. The poor visible transmittance of the films as-deposited without oxygen was increased from ∼ 12% to over 80% on introducing oxygen (3.5 O₂ vol.%). For the films annealed in open air, the average visible transmittance in the wavelength ranging 400-800 nm was varied between 77 and 84%. The films annealed at high temperatures (> 300 °C) decreased the transmittance to as low as < 1%. The optical band gap of the as-deposited films increased from the range 3.83-3.90 to 3.85-3.98 eV on annealing at different conditions.     

Adaptive travelling wave protection algorithm using two correlationfunctions

In this paper, the performance of the correlation function-based travelling wave protection algorithm for high resistance faults is investigated. The reason for its inaccuracies is revealed by the lattice diagram. To eliminate its limitation, an auxiliary cross-correlation function (CCF) is proposed to act as its complement, which takes advantage of the signal that causes the inaccuracies of the standard CCF. For high resistance faults, both the template and the desired signal of the auxiliary CCF have the same prominent feature, which makes them easy to be identified, and the fault location can be acquired from them correctly. An adaptive strategy is designed to select the correct result from the two CCFs so that it can deal with both low resistance and high resistance faults. Simulation results show that the reliability and accuracy of the travelling wave protection have improved significantly     

Poly(L‐lactic acid) metal organic framework composites: optical,thermal and mechanical properties

Biodegradable composites based on poly(L ‐lactic acid) (PLLA) and metal organic frameworks (MOFs) were developed. PLLA without and with the addition of 1, 3 and 5 wt% MOFs was melt compounded in a microextruder. The optical, physical, thermal, mechanical and thermomechanical properties of the composites were evaluated. The Fourier transform infrared, ultraviolet and colorimetric studies showed selected absorption at particular wavelengths due to the presence of copper and benzene belonging to the MOFs. The dynamic mechanical analysis results revealed that the heat deflection temperature, storage modulus and loss modulus of the PLLA–MOF composites did not significantly change compared with the neat PLLA samples. However, a significant decrease in the brittleness of the PLLA–MOF composite was found as evidenced by an increase of 15% in Izod impact strength and 170% in elongation at break. Overall, the brittleness of the PLLA–MOF composite sample decreased as the amount of MOF in the PLLA increased. Copyright © 2011 Society of Chemical Industry     

Potassium-doped sodalite: A tectoaluminosilicate for the catalytic material towards continuous combustion of carbonaceous matters

A novel approach has been proposed for stabilization of potassium in combustion of carbonaceous soot matters emitted from diesel engine vehicles. Sodalite, a clathrate aluminosilicate zeolite, is employed as a support for holding catalytically active potassium species. This approach resulted in a high-catalytic activity and the activity is further enhanced after hydrothermal treatment at 800 °C. Such kinds of tectoaluminosilicate hold potassium strongly and possess high stability against high temperatures and alkaline conditions induced by hydrated potassium ions under practical driving conditions.     

Optimization of ultrasonic welding parameters for copper to copper joints using design of experiments

In this paper, a laser surface micromachining process planning system is presented. In this system, based on a regression model approach, the empirical coefficients, which provide the material removal rate, are automatically generated by a specific software according to the different materials that have to be processed. Numerical models generally present some limits due to the elevated calculation time requested to simulate the laser micromachining of industrial features, especially when transient solutions are considered, and, for this reason, to carry out a useful industrial tool for the evaluation of the material removal rate, the regression model represents the best solution. The presented statistical method, avoiding physical considerations, correlates the material removal rate with the process parameters in a very short calculation time. The automatic procedure for the generation of the coefficients of the regression polynomial permits to easily extend the regression model to any working material and system configuration allowing us to determine the best process parameters in a very short period of time. The results of this work have been patented.     

Solid acid porous materials for the catalytic transformation of 1-adamantanol

Catalytic transformation of 1-adamantanol has been carried out over nano-porous materials with pore sizes ranging from 0.55 to 6 nm and acidic sites ranging from 0.07 to 4 mmol/g. Zeolites viz., H-USY, H-beta, H-mordenite, H-L and H-ZSM-5, and sulfonic-acid functionalized mesoporous silica viz., MCM-41, SBA-15 and SBA-16, were employed for the reaction. Shape selectivity in a confined micropore plays a role with the combination of medium strength acid sites toward formation of desirable 2-derivatives, which comprises a maximum amount of 2-adamantanone.     

Micro-structural and tensile strength analyses on the magnesium matrix composites reinforced with coated carbon fiber

Magnesium matrix composites reinforced with SiO₂ coated carbon fibers have been investigated, with an emphasis given on the relation between the material strength and interfacial microstructure. The composites were studied as a function of aluminium (Al) content that is varied between 0 and 9 wt%. The obtained results indicate that the reactivity at the C/Mg–Al interface of the composite can be controlled by varying the Al content. The low Al content in C/Mg–1Al has been completely dissolved in the matrix with no segregation even after solidification, leading to the best mechanical performance. If the Al content is increased to ≥3 wt% (composites such as C/AZ31 and C/AZ91), the SiO₂ coatings are fully depleted due to an extensive formation of carbides at the interface. The precipitates are further identified as Al₂MgC₂ phase that is similar to binary carbide Al₄C₃. SiO₂ coating on the fiber layer prior to fabrication of composite is found to be a promising way to suppress the carbide formation and enable the use of Mg–Al matrix with appropriate Al content.