A new training strategy for neural network using shuffled frog-leaping algorithm and application to channel equalization

This paper makes use of shuffled frog-leaping algorithm (SFLA) as a training algorithm to train multi-layer artificial neural network (ANN). Next, The SFLA ANNs are used for channel equalization. We, in this paper, also introduce SFLA for channel equalization that is formulated as an optimization problem. In short, this paper introduces a novel strategy for training of ANN and also proposes two novel approaches for channel equalization problem using shuffled frog-leaping algorithm (SFLA). The proposed strategies are tested both in time-invariant and time varying channels and interestingly yield better performance than contemporary approaches as evidenced by simulation results.     

Dual role of betel leaf extract on thyroid function in male mice

The effects of betel leaf extract (0.10, 0.40, 0.80 and 2.0 g kg-1 day-1 for 15 days) on the alterations in thyroid hormone concentrations. lipid peroxidation (LPO) and on the activities of superoxide dismutase (SOD) and catalase (CAT) were investigated in male Swiss mice. Administration of betel leaf extract exhibited a dual role, depending on the different doses. While the lowest dose decreased thyroxine (T4) and increased serum triiodothyronine (T3) concentrations, reverse effects were observed at two higher doses. Higher doses also increased LPO with a concomitant decrease in SOD and CAT activities. However, with the lowest dose most of these effects were reversed. These findings suggest that betel leaf can be both stimulatory and inhibitory to thyroid function, particularly for T3 generation and lipid peroxidation in male mice, depending on the amount consumed.     

Low temperature preparation of nanocrystalline solid solution of strontium barium niobate by chemical process

Sr_xBa_1-xNb₂O₆ (with x = 0.4, 0.5 and 0.6) powders have been prepared by thermolysis of aqueous precursor solutions consisting of triethanolamine (TEA), niobium tartarate and, EDTA complexes of strontium and barium ions. Complete evaporation of the precursor solution by heating at ∼ 200°C, yields in a fluffy, mesoporous carbon rich precursor material, which on calcination at 750°C/2 h has resulted in the pure SBN powders. The crystallite and average particle sizes are found to be around 15 nm and 20 nm, respectively.     

Direct method of overmodulation with integrated closed loop stator flux vector control

Overmodulation achieves full utilization of installed dc link voltage, by modifying the magnitude and the phase of reference voltage vector. Due to the nonlinear behavior in this region, closed loop vector control becomes difficult. A simple method of overmodulation with closed loop control of stator flux vector is proposed in this paper. Instead of modifying the voltage reference by nonlinear equations or look up table, we propose a direct method to generate the switching times. In so doing, the average angular velocity of the stator flux vector is controlled linearly throughout the modulation range. The proposed method shows lower harmonic distortion compared with existing methods. It greatly reduces the complexity and cost of achieving overmodulation over existing schemes.     

Decreased nitric-oxide synthase activity causes impaired endothelium-dependent relaxation in the postischemic heart

Endothelial nitric-oxide synthase (eNOS) is an important regulator of endothelial function and vascular tone in biological tissues. While endothelial dysfunction occurs following ischemia and has been attributed to altered NO. formation, the biochemical basis for this dysfunction is unknown. Therefore, studies were performed to determine the effects of myocardial ischemia and reperfusion on eNOS in isolated rat hearts subjected to periods of global ischemia or ischemia followed by reperfusion. eNOS activity was assayed by L-[14C]arginine to L-[14C]citrulline conversion and alterations in the amount and distribution of eNOS determined by Western blotting and immunohistochemistry. While activity was preserved after 30 min of ischemia with a value of 1.1 +/- 0.1 pmol x min-1 x mg of protein-1, it decreased by 77% after 60 min and became nearly undetectable after 120 min. Reperfusion resulted in only a partial restoration of activity. The decline in activity with ischemia was due, in part, to a loss of eNOS protein. Hemodynamic studies showed that the onset of impaired vascular reactivity paralleled the loss of functional eNOS. Subjecting isolated eNOS to conditions of acidosis, which occur during ischemia, followed by restoration of pH as occurs on reperfusion, caused a combination of reversible and irreversible loss of activity similar to that seen in ischemic and reperfused hearts. Thus, loss of endothelial function following ischemia is paralleled by a loss of eNOS activity due to a combination of pH-dependent denaturation and proteolysis.     

Effect of chemical modification on FTIR spectra. I. Physical and chemical behavior of coir

Coir fibers were chemically modified through alkali treatment, and crosslinking with formaldehyde, para-phenylene diamine, phthalic anhydride, and combined crosslinking-cyanoethylation reactions in appropriate solvent and catalyst. The parent and chemically modified coir were characterized by FTIR spectra. The percent moisture regain, tensile strength, and behavior toward some chemical reagents (solubility %) of parent and chemically modified fibers have also been evaluated. The modified fibers showed significant hydrophobicity, improved tensile strength, and good chemical resistance. © 1995 John Wiley & Sons, Inc.     

Genetic analysis of adult-onset cataract in a city-based ophthalmic hospital

When oxygen delivery (DO2) critically decreases, oxygen consumption (VO2) becomes supply dependent. We examined whether end-tidal PCO2 (PetCO2) would identify supply dependency during shock. Five dogs (Group I) underwent progressive hemorrhage to decrease DO2 until they could no longer maintain a stable blood pressure. Five additional animals (Group II) were bled until VO2 decreased to 70% of baseline, followed by resuscitation. The PetCO2 versus time inflection point was compared with the DO2 at onset of supply dependency (DO2crit). DO2crit for Groups I and II were 6.9 +/- .4 and 8.1 +/- 1.3, respectively (p = NS), and not statistically different from the DO2 values at which PetCO2 decreased (6.6 +/- .7 and 6.3 +/- .7 mL/kg per min, respectively). AT constant minute volume, PetCO2 effectively indicated the onset of supply dependency and rapidly increased during resuscitation, paralleling the changes in VO2 in this model of hemorrhagic shock.     

Synthesis of cyclic carbonates from carbon dioxide and epoxides using alkali metal halide supported liquid phase catalyst

Using combination of Mn–Co transition metal species with N-hydroxyphthalimide as a catalyst for one-step oxidation of cyclohexane with molecular oxygen in acetic acid at 353 K can give more than 95% selectivity towards oxygenated products with adipic acid as a major product at a high conversion (ca. 78%). A turnover number of 74 for this partial oxidation are also recorded.     

Speed ripple minimization in PM synchronous motor using iterative learning control

Permanent-magnet synchronous motor (PMSM) drives are widely used for high-performance industrial servo applications where torque smoothness is an essential requirement. However, one disadvantage of PMSM is parasitic torque pulsations, which induce speed oscillation that deteriorates the drive performance particularly at low-speeds. To suppress these speed ripples, two iterative learning control (ILC) schemes implemented in the time domain and frequency domain respectively are proposed in this paper. Although a conventional proportional-integral (PI) speed controller does suppress speed ripples to a certain extent, it is not adequate for many high performance applications. Thus, the proposed plug-in ILC controller is applied in conjunction with a PI speed controller to further reduce the periodic speed ripples. Experimental verification of the two schemes is carried out, and test results obtained demonstrate that the scheme implemented in frequency domain has better performance in reducing speed ripples than that implemented in time domain because of the elimination of forgetting factor that is indispensable for robustness in time domain learning method.