126,543 research outputs found
Mednarodno trženje električnih ročnih orodij : diplomsko delo visokošolskega strokovnega študija
No full text
Komuniciranje s podrejenimi v bolnišnici Nova Gorica : diplomsko delo univerzitetnega študija
No full text
COD REMOVAL FROM PRE-TREATED EMULSIONS USING ADSORPTION
No full textV okviru diplomske naloge smo čistili odpadne oljno vodne emulzije (OOE) in odpadne oljno vodne emulzije, ki so bile predčiščene z ultrafiltracijo ali z uparjanjem. Cilj naloge je bil odstraniti organske snovi iz emulzij do take mere, da bo kakovost pod predpisano mejno vrednostjo, v skladu z veljavno slovensko zakonodajo. Vzorce smo čistili z adsorpcijo na aktivnem oglju (PAC). Pri čiščenju emulzij, ki niso bile predhodno predčiščene, nismo dosegli želenega znižanja KPK, saj gre za kompleksne emulzije, katerih čiščenje je težavno in zahteva kombinacijo večih metod. Pri adsorpciji predhodno ultrafiltriranih emulzij prav tako nismo dosegli želene mejne vrednosti. Rezultati diplomskega dela so pokazali, da lahko samo v odpadnih emulzijah, ki so bile obdelane z uparjanjem pod določenimi pogoji, z adsorpcijo na aktivnem oglju znižamo KPK vrednost pod 120 mg/L O2.The subject of these thesis was cleaning of wasted oil-in-water emulsions (OOE) and wasted oil-water emulsions, which were pre-treated with ultrafiltration or evaporation. The aim of the thesis was removal of organic compounds from emulsions in order to reach water quality in accordance with slovenian legislation for water release into the environment. The samples of emulsions where cleaned with adsorption on activated carbon(PAC). If the emulisons were not pre-treated, we did not achieve the removal of COD below the limit value, because these emulsions are complex and cleaning of such emuslions requires combination of more different methods. We also did not achieve the limit values if adsorption followed ultrafiltration of emulsions. If the combination of evaporation and adsorption of emulsions was chosen, we could remove COD under 120 mg/L O2
Neuroevolution algorithm neat on graphics cards
No full textV magistrski nalogi naslavljamo problem implementacije algoritma NeuroEvolution of Augmenting Topologies (NEAT) za delovanje na grafičnih karticah. Algoritem NEAT je genetski algoritem za učenje razvijajočih nevronskih mrež. Izhaja iz področja nevroevolucije, ki v umetni inteligenci uporablja genetske algoritme za generiranje in učenje nevronskih mrež. Algoritem za svoje delovanje porabi veliko strojnih in časovnih virov, zato je implementacija na grafičnih karticah smiselna. Implementacijo smo izvedli v arhitekturi CUDA, ki jo podpirajo grafične kartice podjetja NVIDIA. Hitrost in uspešnost algoritma smo izmerili na petih različnih grafičnih karticah in jo primerjali s hitrostjo in uspešnostjo originalnega algoritma. Ugotovili smo, da je naša implementacija algoritma zadovoljiva, saj je hitrejša in prav toliko uspešna kot originalna implementacija algoritma NEAT.We address the problem of NeuroEvolution of Augmenting Topologies (NEAT) algorithm implementation for operating on graphics cards. NEAT is a genetic algorithm for learning and evolving neural networks. It’s a member of the neuroevolution algorithms which use genetic algorithms to learn and evolve neural networks in the field of artificial intelligence. Algorithm uses a lot of resources for its operation and is, therefore, suitable for implementation on graphics cards. We implemented it on a CUDA architecture, which is supported by NVIDIA graphics cards. We measured the speed and performance of the algorithm on five different graphics cards and compared it to the speed and performance of the original algorithm. Our implementation is satisfactory because it is faster than and just as efficient as the original NEAT implementation
ZINC AND COPPER REMOVAL FROM WATER USING MODIFIED ACTIVATED CARBON CLOTH
No full textNamen magistrskega dela je bil modificirati vlakna iz aktivnega oglja z naravnim polimerom hitozanom. V nadaljevanju smo izvedli še študijo karakterizacije vlaken ter adsorpcijo kovin na modelnih in realnih kompostnih vodah.
Vlakna smo modificirali z 1 % in 2 % raztopino hitozana. Med seboj smo primerjali nemodificirano referenčno vlakno aktivnega oglja in modificirani vlakni. S študijo karakterizacije smo želeli dokazati, da smo na vlakna uspešno vezali biopolimer hitozan, ki služi kot adsorber kovin. V ta namen smo posneli FTIR in Raman spektre modificiranih vlaken, vendar nismo dobili značilnih vrhov za hitozan. Razen tega smo izvedli potenciometrično titracijo in določili število amino skupin na posameznih vlaknih. Spomočjo polielektrolitske titracije smo spremljali desorpcijo hitozana iz površine vlaken in dokazali, da gre pri modifikaciji za ireverzibilno reakcijo.
Modificirana vlakna smo v nadaljevanju uporabili za adsorpcijo Cu2+ in Zn2+ ionov, pričemer smo s hitrimi testi določili, da je pH= 5 najbolj optimalna vrednost za adsorpcijo izbranih kovinskih ionov. Vsebnost preostalih kovin v kopeli, po adsorpciji na modificirana vlakna smo določali s pomočjo AAS. Rezultati kažejo, da se na vlakna modificirana z 2 % raztopino hitozana, Zn2+ in Cu2+ ioni (pri začetnih koncentracijah < 20 mg/L) vežejo v tolikšni meri, da lahko očiščeno vodo spustimo v okolje. Referenčna nemodificirana vlakna pa Zn2+ in Cu2+ ne odstranijo zadostno. Pri realnih kompostnih vodah smo odstranili do 61,1 % cinka in 100 % bakra. Raziskava je pokazala, da je adsorpcija izbranih kovin na vlakna modificirana s hitozanom uspešnejša v primerjavi z adsorpcijo na referenčna vlakna.The aim of the master’s thesis was to modify the activated carbon fibres using the natural polymer chitosan. Further, the characterisation study of fibres and adsorption of metals from model and real compost water was performed.
Fibres were modified with 1% and 2% chitosan solution. Modified fibres were then compared to unmodified referential activated carbon fibre. The characterisation study was performed in order to show that biopolymer chitosan, which serves as metal adsorbent, was successfully bonded on activated carbon fibres. For this purpose, FTIR and Raman spectra of the modified fibres were recordedhowever, peaks that are typical for chitosan were not observed.
With potentiometric titration, the number of amino groups on individual fibres was determined. By polyelectrolyte titration, desorption of chitosan from a fibre surface was monitored, which showed that irreversible reaction is typical for modification.
Modified fibres were then used for adsorption of metalsand after performing quick tests, pH 5 was determined as the most optimal pH value for adsorption. The residual concentration of metals in water solution after the adsorption by modified fibres was determined with AAS. Results have shown that Zn2+ and Cu2+ were bonded (at initial concentration of < 20 mg/L) on activated carbon fibres modified with 2% chitosan solution in such extent, that treated model water can be released into the environment. Referential unmodified fibres do not remove Zn2+ and Cu2+ adequately. Otherwise, from real compost water up to 61,1 % of zinc and 100 % of copper was eliminated. However, the research has shown that adsorption of chosen metal ions on fibres modified with chitosan was more sucesful than adsorption on referential unmodified fibres
PROGRAM INTERFACE FOR DYNAMIC GESTURE RECOGNITION WITH MYO ARMBAND
No full textV diplomskem delu je predstavljen programski vmesnik, ki bo znal razpoznavati geste roke pri interakciji človek – stroj. Tehnologija prehaja v obdobje, kjer se uveljavljajo novi načini interakcije človeka s strojem in nam omogoča lažje upravljanje in učenje strojev samih. Čeprav še vedno na področju interakcije s stroji prevladujeta miška in tipkovnica, se v ospredje prebijajo tudi druge naprave, ki omogočajo enake ali celo boljše interakcije človek – stroj. Eden izmed takšnih proizvodov je tudi zapestnica Myo, ki je v diplomskem delu uporabljena za zajemanje in analizo poljubnih uporabnikovih gest roke. Napisan je programski vmesnik, ki je zmožen učenja in razpoznavanja gest, ter testirano, kako dobro se obnese pri razpoznavanju različnih gest. Meritve nakazujejo, da je programski vmesnik pravilno razpoznal večino naučenih gest, iz česar je možno sklepati, da je zmožen prepoznati dinamično naučene geste.We describe a program interface for dynamic gesture recognition in human-computer interaction. Technology is forcing new ways of human-machine interaction which enables faster learning and more complex controlling of machines, such as computers. Mouse and keyboard are still dominant devices for interaction with computers, but other devices are rushing ahead. One of these devices is Myo armband, which has been used in our work to acquire and analyze the hand gestures. We wrote a program interface, which learns arbitrary gestures and systematically analyses differences between different gestures in order to illuminate the accuracy of their automatic classification and discrimination. We tested the interface on 5 gestures and demonstrated that our program recognized most of learnt gestures and we can say it’s capable of recognizing dynamically learnt gesture
For girls (above) and boys (below): a) Observed growth for 15 randomly selected children from every cohort, b) SITAR-based cohort-specific average growth curves with height measurements, and c) SITAR-based growth velocity curves.
No full textFor girls (above) and boys (below): a) Observed growth for 15 randomly selected children from every cohort, b) SITAR-based cohort-specific average growth curves with height measurements, and c) SITAR-based growth velocity curves.</p
- …
