Doronkin D.E.,Karlsruhe Institute of Technology |
Kuriganova A.B.,Platov South Russian State Polytechnical University |
Leontyev I.N.,Southern Federal University |
Baier S.,Karlsruhe Institute of Technology |
And 3 more authors.
Catalysis Letters | Year: 2016
Pt/γ-Al2O3 catalysts made by fast and simple electrochemical dispersion method were characterized using X-ray absorption spectroscopy, CO chemisorption, transmission electron microscopy and X-ray diffraction, and compared with an impregnated catalyst with respect to oxidation of CO and NO. A combination of techniques revealed average particle sizes of 3-4 nm for 0.81-3.8 wt% Pt/γ-Al2O3 catalysts. Electrochemically prepared materials demonstrated catalytic activity comparable to that of conventional impregnated catalyst and reasonable stability. © 2015 Springer Science+Business Media New York.
Bogush I.A.,Platov South Russian State Polytechnical University |
Cherkashin V.I.,Russian Academy of Sciences |
Ryabov G.V.,Platov South Russian State Polytechnical University |
Abdullayev M.S.,Russian Academy of Sciences
Doklady Earth Sciences | Year: 2016
The weathering crust of the Beden ultrabasite massif (the basin of Big Laba River) is identified and studied. Anomalously high contents of noble metals (Au, Pt, Pd) are revealed in the basal horizon of the Jurassic part of the weathering crust. For this reason we suspect an existence of a belt of noble metal miner-alization in the Paleozoic ultrabasites in the Peredovoi Range of the Northern Caucasus. © 2016, Pleiades Publishing, Ltd.
Dubrov V.I.,Platov South Russian State Polytechnical University |
Kirievskiy E.V.,Platov South Russian State Polytechnical University |
Vladimirovna Savchenko A.,Platov South Russian State Polytechnical University
Life Science Journal | Year: 2013
This paper describes the diagnostics algorithms of one of the most important elements of the electric mains - the circuit breaker. There are the descriptions of two versions of the intelligent diagnostic algorithm of highvoltage circuit breakers based on neural networks. The main attention is paid to the comparison of the described algorithms on the spent resources, both in time and in computation. The article also describes the final stage of development of the diagnostic algorithm which is to optimize such parameters of the neural network as the number of neurons in a layer, the number of hidden layers in the neural network, the parameter of the network learning rate, the number of epochs (iterations) for training in the network. To test the adequacy of the neural network with optimized parameters we used the method of cross-validation. The algorithms are simulated in the Matlab environment, and the researchers have chosen high-voltage switch of MKP 110M type as the diagnostics object.