Fraunhofer Chalmers Research Centre for Industrial Mathematics
Chalmers Publication LibraryNot a member yet
70439 research outputs found
Sort by
Spin Dependent Discrimination between Majorana and Dirac Dark Matter
No full textA large number of observations reveal deviations from expected gravitational behaviour
in astrophysical systems. This suggests that something is missing in our
understanding of the Universe. A well established candidate for explaining these
deviations is dark matter, a kind of matter that is not subject to photon interactions
and thus can not be detected by absorbed or emitted light - it is dark. Even
though we can not observe it at any detectable wavelength, dark matter is postulated
to interact via gravitation. These interactions would then account for the
shortcoming of gravitational pull from visible mass, explaining the observed irregular
gravitational behaviour. One leading hypothesis is that the proposed dark matter
component of the Universe is actually made out of massive, weakly interacting particles.
As of today, there has not yet been any detection of such a particle, leaving
the fundamental properties of dark matter unknown.
There are currently several ongoing experimental projects searching for dark matter
particles, and more are being planned. Furthermore, their accuracy is higher
than ever, and if the dark matter particle hypothesis is correct, it is reasonable
to expect a detection signal in the foreseeable future. The different searches take
complementary approaches and exploit either annihilation, production or direct detection
experiments. The latter, which is the focus of this thesis, aims to detect
dark matter by measuring the recoil of a target nucleus in a detector when a dark
matter particle scatters off of it. In case of positive detection at direct detection
experiments, a model describing the interaction between dark matter particles and
baryonic matter is needed in order to be able to draw conclusions about the dark
matter particle properties.
Under the assumption that dark matter is a spin 1/2 particle that only interacts
with baryonic matter via spin dependent interactions, this thesis studies dark
matter-nucleus scattering in order to see if it is feasible to discriminate between
Majorana and Dirac dark matter. I find that, if dark matter particles are detected
at three different experiments, the Majorana dark matter hypothesis can be rejected
in favour of an alternative hypothesis in which dark matter is a Dirac particle. Restrictions
on target elements that are of interest for this test are presented, and the
test procedure is studied for setups containing some of them, namely: 131Xe, 127I,
73Ge, 23Na and 19F
Strategier för aerodynamisk formoptimering av bakvingen på en generisk tävlingsbil
No full textFormoptimering inom fordonsaerodynamik har traditionellt sett skett genom trial
and error, eller med hjälp av tester i vindtunnlar. Inom GT3-racing är det viktigt
att få så mycket negativ lyftkraft som möjligt utan att öka luftmotståndet för mycket.
Med datorers ökade beräkningskapacitet har numeriska simuleringar blivit allt
mer vanligt och därför finns det intresse att undersöka hur olika optimeringsmetoder
lämpar sig för fordonsaerodynamik. Denna rapport syftar till att utvärdera
tre optimeringsmetoder, på den inom forskning väl använda DrivAer-modellen, för
att optimera position och attackvinkel på en bakvinge. De optimeringsmetoder som
undersökts i projektet är surrogatmodellering, genetiska algoritmer och gradientbaserad
formoptimering av adjointa fält.
Surrogatmodellen genomfördes med två optimeringsvariabler, höjd och attackvinkel,
för att senare i projektet kompletteras med en tredje variabel, vingens position i longitudinellt
led. Med ett få antal simuleringar konvergerade höjden och attackvinkel
mot optimala värden för att minska luftmotståndskoefficienten, CD, med en bibehållen
lyftkoefficient, CL. Resultaten visar att modellen optimerar två variabler snabbt
trots en stor lösningsrymd.
Den genetiska algoritmen har visat sig vara mycket beräkningstung, då det krävs
många generationer för att med säkerhet uppnå ett optimalt resultat. Metoden
verkade konvergera mot ett visst värde. Studien är dock för kort för att veta om
det var ett globalt minimum. Den genetiska algoritmen faller kort jämfört med
surrogatmodellen då utvärderingen av varje individ är för beräkningstung för att
göra den genetiska algoritmen effektiv.
Den gradientbaserade formoptimeringen med adjointa fält skiljer sig i sitt utförande
jämfört med surrogatmodellen och den genetiska algoritmen. Dessa metoder optimerar
utifrån variabler medan adjointmetoden istället optimerar genom att förändra
geometrin. Adjointmetoden visade sig ha svårt att konvergera för detta fall. Optimering
med en stegfunktion fungerade inte, men användandet av en sigmoid funktion
var mer lovande. För att utnyttja adjointmetoden effektivt ansågs det vara bra att
använda surrogatmodellering eller genetiska algoritmer tillsammans med adjointmetoden.
Nyckelord
Lane-Level Map Matching using Hidden Markov Models
No full textMap matching is the procedure of matching vehicle location and sensor
data to a digital map. New high-definition maps, designed for autonomous
vehicles, open up for the possibility of matching to lanes rather than roads.
Inferring the lane-level positions of vehicles will be useful for updating
and building probe-sourced maps, and thereby arguably essential for autonomous
driving.
This thesis seeks to solve the lane-level map matching problem using a
Hidden Markov Model. The Viterbi algorithm is used to decode it. The
model is tested on a data set yielded through the Volvo Drive Me project
and collected by commercial vehicle sensors, including a GPS receiver, an
inertial navigation system and a forward-looking camera. For the sake of
simplicity, the RADAR and LiDAR sensors are excluded. Among the sensor
data used, lane changes and the type of road lane markings as detected by
the vehicle proves to be particularly important.
Two metrics for evaluating model performance are proposed. The first
metric is the recall, i.e. the fraction of correct matches. However, the lanes
to which the observations are matched vary widely in length. Therefore, we
introduce the path length error (PLE) as a complementary metric. As the
name indicates, it considers the length of the incorrect routes.
A naive matcher, that simply matches GPS coordinates to the closest lane,
is used for benchmarking. Attaining 95% median recall and 3% median
PLE, we conclude that our model is high-performing and robust to errors.
For comparison, the naive matcher scores 77% median recall and 26% median
PLE. Our model is however shown to struggle without reliable vision
detections. It would therefore be meaningful to investigate the inclusion of
additional vehicle sensors
Electronic Stability Control Assessment Methods: A Study in the Correlation Between Subjective Assessment and Objective Measurement in Vehicle Testing
No full textThis report describes the work to include objective measurement methods in real car electronic stability control testing, conducted at CEVT vehicle dynamics department. Current testing methodology is focused on subjective assessment by the driver, with no objective way to judge test data. By post processing logged test data from single and double lane change maneuvers and correlating it to the subjective assessment of the test driver, objective rating methods were established. To gather reliable data, dedicated vehicle testing of both Lynk & Co and competitor cars was carried out at Volvo Hällered Proving Ground by expert drivers. To measure the performance of competitor cars, a new measurement equipment set-up was created for benchmark testing, consisting of steering effort sensor and inertial measurement unit.
As part of the project, a simulation study of single lane change was also carried out, to enhance understanding of the vehicle’s behavior and to bridge the gap between simulation and real car testing. The simulation drive case was also tested in the real vehicle.
The result of these processes is rating limits on side slip, yaw rate, longitudinal jerk and longitudinal velocity delta for both single and double lane change. When verified these limits can be the foundation for objective requirements on single and double lane change performance
Realtidsövervakning på järnvägen - En förstudie om implementering av QTMS på Öresundsförbindelsen
No full textEFFECT OF MONOMERIC AMYLOID-β ON CELL SECRETED EXOSOMES Discerning the pathology of Alzheimer’s disease
No full textAmyloid-β is a significant protein associated with the pathogenesis of Alzheimer’s disease. Dysregulation and overproduction of the protein after cleavage from the transmembrane Amyloid Precursor Protein (APP) by β and γ secretases, will lead to its aggregation into amyloid fibrils and subsequent accumulation into senile plaques. The development and progression of plaque pathology across the brain is gradual and Alzheimer’s patients will progressively show marked cognitive disabilities, decreased social abilities, dementia and reduced life expectancy. In healthy individuals, there are several clearance mechanisms that control the levels of Aβ42 in the body and prevent its aggregation. They include several intracellular processes that maintain cellular homeostasis via autophagy systems and microglial intervention. Exosomes are a class of cell secreted nanovesicles that are involved in many cell signaling processes as well as spread of neurological disorders such as Parkinson’s disease, ALS and Alzheimer’s disease. They are formed through exocytosis of intraluminal vesicles of multivesicular bodies, a trafficking organelle in the cell. Exosomes trap many proteins for cell signaling, transport, and intercellular communication. There is literature that describes the spread of Alzheimer’s disease through the formation of misfolded and aggregated forms of Aβ42, including their association with exosomes, which could mediate release into the extracellular environment as well as propagation to neighboring cells. But nothing is known about how Aβ42 peptides affect the exosome population that is secreted from a neuronal cell. Can Aβ accumulations in neurons alter exosome numbers, and alter exosome size distributions, which in turn could lead to release of more Aβ? How does Aβ peptides interact with the exosome?
The aim of my thesis is to explore the effect of Aβ peptides on the exosome population secreted from a human neuroblastoma (SH-SY5Y) cell model. To achieve this, I developed small volume protocols to treat SHSY5Y cells with Aβ monomers and to isolate extracellular vesicles from harvested media. The extracellular vesicles were analysed for exosome markers and Aβ content using Western blot, their population size and numbers were determined by nanoparticle tracking analysis (NTA), and thioflavinT (ThT) aggregation assays were used to monitor the ability of exosomes to catalyze the in vitro aggregation of Aβ monomers. Further, using a modified SH-SY5Y cell line, expressing a GFP-tagged version of the exosome marker CD63, I used time lapse confocal microscopy to analyse the in-situ colocalization of Aβ with exosomes. Also, to scale down the experiments and analyze the exosome particles released in real time, we are developing a micro/nanofluidic device that will contain a live culture of SHSY5Y cells to analyze exosomes released by them through nanochannels built in it. Through these experiments, I found that exosomes secreted from cells that had been treated with monomeric Aβ42 had higher concentration and also increased mean size (diameter) compared to untreated control (mock). The exosomes slowed down the aggregation of Aβ42 into fibrils, primarily by extending the lag phase of the aggregation reaction. Altogether, these results suggest that exosomes are not only carriers of Aβ42 peptides, but that their physicochemical properties and concentrations can also be altered, which in turn can have both aggravating and protective effects on the accumulations of Aβ42 that occur during the development and progression of Alzheimer’s disease
Produktutveckling av fixeringsblock
No full textProjektet syftade till att reducera kostnaden av en komponent i en produkt. Detta skulle utföras så kvaliteten inte påverkades negativt. Idag tillverkas produkten genom att fräsa ut komponenten ur ett aluminiumblock. Företaget har också tittat på prototyper i plastkompositer men plastkompositen de har kollat på har visat sig vara för vek trots att den hade en E-modul
på nästan 14 GPa. Genom ett produktutvecklingsarbete har flera koncept utvecklats och därefter sållats. De
koncept som rapporten har kommit fram till har en god utvecklings- och tillverkningspotential. Resultatet visade att pressgjuta i zink skulle leda till en
kostnadsreducering som dessutom bibehåller kvaliteten. Fördelen med zink är att det är ett material som är lätt att gjuta i, miljövänligt och den har högre E-modul än aluminium. De valda koncepten har analyserats noga med FEM-beräkningar och dessa har visat att koncepten
klarar av de kriterier som har ställts.
Utöver konceptgenerering så har projektet behandlat områden som hur man sänker tillverkningskostnaden på den befintliga modellen, materialval, formändringar och även hur mycket man kan behöva kompromissa för att ett koncept ska vara bättre än ett annat