The Combined treatment of UV Light and Titanium Dioxide (TiO2) in Grease Filtration Technique

The aim of this thesis was to identify the best grease removal technique with the application of low power of UV light to TiO2 coated grease filters. The treatment with various power series of ozone generating and ozone free lamps to normal grease filters and TiO2 coated grease filters were examined and the obtained results are compared to each other in this paper. The effect of ozone reaction was observed and compared with the effect of TiO2. The experiments were solely based on the photo oxidation and photo catalytic oxidation reactions. TiO2 is a green catalyst used in the photocatalytic reaction. Sunflower oil was used for grease production and tetracholoroethylene as a solvent. Grease samples were collected from the ventilation duct connected to the cooking hood system. Sample extraction was done in ultrasonic bath with the principle of sonication. The sample analysis was done by FTIR machine. The result determining the concentration of grease was the quantification of saturated C-H bonds in the chosen peak group of the spectrum. A very low power of UVC light functions perfectly with the Titanium dioxide. The experimental results have shown the combined treatment of titanium dioxide and UV light is an effective method in grease removal process. The photocatalytic reaction with titanium dioxide is better than photo oxidation reaction with ozone treatment. Photocatalytic reaction is environmentally friendly, energy efficient and economical.

Transient current technique simulation of semiconductor detectors

Nowadays advanced simulation technologies of semiconductor devices occupies an important place in microelectronics production process. Simulation helps to understand devices internal processes physics, detect new effects and find directions for optimization. Computer calculation reduces manufacturing costs and time. Modern simulation suits such as Silcaco TCAD allow simulating not only individual semiconductor structures, but also these structures in the circuit. For that purpose TCAD include MixedMode tool. That tool can simulate circuits using compact circuit models including semiconductor structures with their physical models. In this work, MixedMode is used for simulating transient current technique setup, which include detector and supporting electrical circuit. This technique was developed by RD39 collaboration project for investigation radiation detectors radiation hard properties.

Testing FCM technique for the wind energy scenarios in Finland

In the latter days, human activities constantly increase greenhouse gases emissions in the atmosphere, which has a direct impact on a global climate warming. Finland as European Union member, developed national structural plan to promote renewable energy generation, pursuing the aspects of Directive 2009/28/EC and put it on the sharepoint. Finland is on a way of enhancing national security of energy supply, increasing diversity of the energy mix. There are plenty significant objectives to develop onshore and offshore wind energy generation in country for a next few decades, as well as another renewable energy sources. To predict the future changes, there are a lot of scenario methods developed and adapted to energy industry. The Master’s thesis explored “Fuzzy cognitive maps” approach in scenarios developing, which captures expert’s knowledge in a graphical manner and using these captures for a raw scenarios testing and refinement. There were prospects of Finnish wind energy development for the year of 2030 considered, with aid of FCM technique. Five positive raw scenarios were developed and three of them tested against integrated expert’s map of knowledge, using graphical simulation. The study provides robust scenarios out of the preliminary defined, as outcome, assuming the impact of results, taken after simulation. The thesis was conducted in such way, that there will be possibilities to use existing knowledge captures from expert panel, to test and deploy different sets of scenarios regarding to Finnish wind energy development.

Sentence Compression with Supervised Machine Learning

The subject of the thesis is automatic sentence compression with machine learning, so that the compressed sentences remain both grammatical and retain their essential meaning. There are multiple possible uses for the compression of natural language sentences. In this thesis the focus is generation of television program subtitles, which often are compressed version of the original script of the program. The main part of the thesis consists of machine learning experiments for automatic sentence compression using different approaches to the problem. The machine learning methods used for this work are linear-chain conditional random fields and support vector machines. Also we take a look which automatic text analysis methods provide useful features for the task. The data used for machine learning is supplied by Lingsoft Inc. and consists of subtitles in both compressed an uncompressed form. The models are compared to a baseline system and comparisons are made both automatically and also using human evaluation, because of the potentially subjective nature of the output. The best result is achieved using a CRF - sequence classification using a rich feature set. All text analysis methods help classification and most useful method is morphological analysis. Tutkielman aihe on suomenkielisten lauseiden automaattinen tiivistäminen koneellisesti, niin että lyhennetyt lauseet säilyttävät olennaisen informaationsa ja pysyvät kieliopillisina. Luonnollisen kielen lauseiden tiivistämiselle on monta käyttötarkoitusta, mutta tässä tutkielmassa aihetta lähestytään television ohjelmien tekstittämisen kautta, johon käytännössä kuuluu alkuperäisen tekstin lyhentäminen televisioruudulle paremmin sopivaksi. Tutkielmassa kokeillaan erilaisia koneoppimismenetelmiä tekstin automaatiseen lyhentämiseen ja tarkastellaan miten hyvin erilaiset luonnollisen kielen analyysimenetelmät tuottavat informaatiota, joka auttaa näitä menetelmiä lyhentämään lauseita. Lisäksi tarkastellaan minkälainen lähestymistapa tuottaa parhaan lopputuloksen. Käytetyt koneoppimismenetelmät ovat tukivektorikone ja lineaarisen sekvenssin mallinen CRF. Koneoppimisen tukena käytetään tekstityksiä niiden eri käsittelyvaiheissa, jotka on saatu Lingsoft OY:ltä. Luotuja malleja vertaillaan Lopulta mallien lopputuloksia evaluoidaan automaattisesti ja koska teksti lopputuksena on jossain määrin subjektiivinen myös ihmisarviointiin perustuen. Vertailukohtana toimii kirjallisuudesta poimittu menetelmä. Tutkielman tuloksena paras lopputulos saadaan aikaan käyttäen CRF sekvenssi-luokittelijaa laajalla piirrejoukolla. Kaikki kokeillut teksin analyysimenetelmät auttavat luokittelussa, joista tärkeimmän panoksen antaa morfologinen analyysi.