This paper serves as a tutorial on the use of neural networks for solving combinatorial optimization problems. It reviews the two main classes of neural network models : the gradient-based neural networks such as the Hopfield network, and the deformable template approaches such as the elastic net method and self organizing maps. In each class, the original model is presented, its limitations discussed, and subsequent developments and extensions are reviewed. Particular emphasis is placed on stochastic and chaotic variations on the neural network models designed to improve the optimization performance. Finally, the performance of these neural network models is compared and discussed relative to other heuristic approaches.
The lowest part of the Monarch Mill Formation in the Middlegate basin, west-central Nevada, has yielded a middle Miocene (Barstovian Land Mammal Age) vertebrate assemblage, the Eastgate local fauna. Paleobotanical evidence from nearby, nearly contemporaneous fossil leaf assemblages indicates that the Middle Miocene vegetation in the area was mixed coniferous and hardwood forest and chaparral-sclerophyllous shrubland, and suggests that the area had been uplifted to 2700–2800 m paleoaltitude before dropping later to near its present elevation of 1600 m. Thus, the local fauna provides a rare glimpse at a medium- to high-altitude vertebrate community in the intermountain western interior of North America. The local fauna includes the remains of fish, amphibians, reptiles, birds, and 25 families of mammals. Carnivorans, the focus of this study, include six taxa (three of which are new) belonging to four families. Canidae are represented by the borophagine Tomarctus brevirostris and the canine Leptocyon sp. indet. The earliest record and second North American occurrence of the simocyonine ailurid Actiocyon is represented by A. parverratis sp. nov. Two new mustelids, Brevimalictis chikasha gen. et sp. nov. and Negodiaetictis rugatrulleum gen. et sp. nov., may represent Galictinae but are of uncertain subfamilial and tribal affinity. The fourth family is represented by the felid Pseudaelurus sp. indet. Tomarctus brevirostris is limited biochronologically to the Barstovian land mammal age and thus is consistent with the age indicated by other members of the Eastgate local fauna as well as by indirect tephrochronological dates previously associated with the Monarch Mill Formation. Actiocyon parverratis sp. nov. extends the temporal range of the genus Actiocyon from late Clarendonian back to the Barstovian. The Eastgate local fauna improves our understanding of mammalian successions and evolution, during and subsequent to the Mid-Miocene Climatic Optimum (~14–17 Ma).
Badano powiększanie skali procesu wytrącania proszków przy użyciu dwutlenku węgla w stanie nadkrytycznym jako substancji zmniejszającej rozpuszczalność. Dokonano analizy procesu pod kątem etapów kluczowych i zaproponowano optymalne kryterium powiększania skali. Wybrane kryterium zweryfikowano doświadczalnie.
EN
Scale up of the supercritical fluid antisolvent precipitation process was investigated. Optimal scale up criterium was suggested based on process characteristics. Chosen criteria were then experimentally verified at laboratory scale, in a pilot plant and in a smal! manufacturing plant.
Badano rozpad kropel etanolu w dwutlenku węgla w stanie nadkrytycznym w dyszy koncentrycznej wykorzystywanej do produkcji proszków metodą wytrącania. W pracy rozpatruje się dwa mechanizmy rozpadu kropel: pierwszy wynika z różnicy pomiędzy prędkościami kropel i fazy ciągłej (na granicy faz powstają niestabilności Rayleigha-Taylora), drugi - z występowania naprężeń generowanych przez fluktuacje burzliwe prędkości płynu. Wyniki obliczeń zostały porównane z wynikami doświadczeń.
EN
Breakup of ethanol droplets in the supercriticaI carbon dioxide by a twin-fluid nozzle is considered in relation to powder production. Two mechanisms of drop breakup are considered: the first resulting from the volocity difference between the drop and the continuous phase (Rayleigh-Taylor instability) and the second one resulting from the stress generated by velocity fluctuations. Model predictions are compared with experimental results.