Thesis work: Development of classification algorithms based on s
Swerea Kimab AB / Sjukgymnastjobb / Stockholm
Observera att sista ansökningsdag har passerat.
Visa alla sjukgymnastjobb i Stockholm,
Solna,
Lidingö,
Sundbyberg,
Danderyd eller i
hela Sverige Visa alla jobb hos Swerea Kimab AB i Stockholm,
Solna eller i
hela Sverige Background
Aluminium is 100% recyclable. Its recycling involves the collection of waste, sorting, melting and subsequent use as secondary material in the production of new products. The use of recycled aluminium requires only 5% of the energy used in extraction of virgin minerals to produce aluminium, thus enabling large savings in energy consumption.
A large part of the aluminium scrap on the market today is delivered from shredding mills, in which cars as well as industrial and household goods are cut into small pieces. The shredded material is inhomogeneous and, today, mostly sorted by visual inspection or coarse sorting techniques. The uncertainties in alloy composition of scrap materials set an upper bound to the amount of recycled aluminium used in production, where very stringent compositions are required.
In recent years, laser-induced breakdown spectroscopy (LIBS) has been established as a promising analytical tool for online chemical analysis. The emitted light spectrum is analysed for instantaneous determination of the elemental composition of the sample, enabling on-line classification of materials. Swerea KIMAB has, in collaboration with RISE Acreo, developed prototypes of LIBS systems designed for the on-line analysis and sorting of scrap flows. Experimental field trials at our partners' sites in Sweden, China and Brazil have provided very good results and indicate that alloy classification and quantitative analysis can be implemented on unprepared scrap aluminium samples. However, most on-line classification algorithms are today based on the use of a limited number of features extracted from the spectral data (so-called optical emission lines). In order to fully exploit the advantages of the LIBS technology, more advanced classification techniques based on spectral data could be further developed and improved (use of decision trees, artificial neural networks, multivariate analysis, chemometrics, etc.).
Thesis description
• Literature study and choice of the most promising types of classifiers.
• Implementation of classification techniques in Matlab.
• Evaluation of classification techniques on available training and test sets.
• Further development, optimisation and final assessment.
• Report.Requirements
• Knowledge in data analysis/machine learning.
• Experience in Matlab programming.
• Substantial interest in developing, implementing and evaluating data analysis techniques.
• No prior knowledge in LIBS or other measurement techniques is required. Interest in spectroscopy is however beneficial.
Applications should include a brief personal letter, CV, and recent grades.
Location
Swerea KIMAB, Kista (Stockholm)
Contact persons
Bertrand Noharet (
bertrand.noharet@swerea.se)
Tania Irebo Schwarz (
tania.irebo@swerea.se)
Application and time
Application is made through our website (www.swereakimab.se/career). We want you to submit a personal letter, CV and recent grades no later than April 20, 2017. The diploma work is to be started at the latest in May 2017.
About Swerea KIMAB
We are an institute in materials science, with leading-edge skills in materials development, processing and analysis, and unique expertise in corrosion and corrosion protection. Most of our revenue comes from commissioned industrial research, but we are also very actively involved in various national and European research programmes. We are contributing to reinforcing the competitive strength of our clients and to developing new materials and product solutions within areas such as transport, energy, environment and infrastructure. At our new laboratories in Stockholm we provide testbeds and demonstrators for developing and verifying production processes for optimized material performance. We have laboratories for modelling and simulation of manufacturing processes for advanced metallic materials. Process analytical monitoring is another area in which we have invested considerable effort in recent years.
Varaktighet, arbetstid
Heltid
Publiceringsdatum2017-02-13ErsättningMånatlig löneutbetalning
Så ansöker duSista dag att ansöka är 2017-04-20
Klicka på denna länk för att göra din ansökanFöretagSwerea KIMAB AB
AdressSwerea KIMAB AB
Box 55970
10216 Stockholm
KontorsadressBox 55970 10216 Stockholm
Jobbnummer 3316549
Observera att sista ansökningsdag har passerat.