Fraunhofer Chalmers Research Centre for Industrial Mathematics
Chalmers Publication LibraryNot a member yet
70439 research outputs found
Sort by
Condition monitoring system for wind turbines – based on deep autoencoders
Wind turbines consist of many mechanical, electrical and hydraulic com-
ponents. Failures in any of these can lead to high financial loss, both as actual repair costs
and from lost production time. Often, failures do not occur instantaneously but rather as a
consequence of sequential degrad- ing. Therefore, many failures can be detected in advance using
so-called condition monitoring systems. Through their supervisory control and data acquisition
system, modern wind turbines store information about their op- erating state. Among other things,
signals such as produced power, wind speed, and various component temperatures are recorded. In
this thesis, a condition monitoring system that leverages this data is developed. The system is
based on deep autoencoders, a type of neural network that learns to reconstruct its input data. By
training an autoencoder on data from a healthy wind turbine it can learn the dependencies between
different SCADA signals under normal conditions. If it then gives a poor reconstruction for new
data, it is likely that something has changed in the internal dynamics of the wind turbine which
could indicate a degraded component.
Previously, many similar systems have been developed. These have shown good results and detected
multiple component failures up to months in advance. However, they usually only monitor one
component at a time and are therefore not able to provide a complete condition monitoring sys-
tem. Autoencoders, which do not suffer this problem, have also been inves- tigated but not at a
larger scale.
In this thesis, a relatively large, labeled dataset was utilized. With this data, the efficiency of
condition monitoring systems based on autoencoders was tested on a variety of real faults.
Moreover, the influence of various prop- erties of the autoencoder was investigated. The results of
the investigation showed that an autoencoder based condition monitoring system is capable of
detecting a variety of failures in wind turbines. Finally, suggestions for future developments are
discussed in the thesis
Innovations in Finance as Regulators Push Open Banking - Understanding innovative processes and market change in the Swedish banking market as of payment services directive 2
Digital technologies are transforming each layer of every stack in society. Within the financial
industry, authorities in the EU and in Sweden believe that the development of digital technologies
is being held back. Competitive authorities urge for the industry to provide better products,
increased customer choice and higher price competition. One regulation that aims toward
enhancing innovation within banking in Europe is the directive payment services directive 2
(PSD2), introduced in 2018. PSD2 requires banks to enable external parties to use their account
services, and transaction services, through technical APIs. Banks’ compliance of PSD2 is
sometimes referred to as open banking.
The purpose of this report is to, from an academic perspective, explore the market changes in
Sweden as of PSD2 and open banking. Specific theory related to technologies, and more general
theory applicable on market change, disruption, and innovation, is used. Further, disruptive
tendencies toward incumbent banks are analyzed.
The study shows that there are reasons to believe that the industry has moved to an un-optimal
state in terms of innovation, and that PSD2 can partly move it to a more efficient state. It shows
strong incentives for collaborations between FinTechs and banks post PSD2 and open banking,
based on utilizing each other’s strengths to enhance value offerings toward end consumers. The
research concluded that there are no direct disruptive effects of PSD2 and open banking. However,
PSD2 and open banking lowers the efforts for a business to configure itself as a marketplace bank.
While incumbent banks rely on providing their own provision services and loans on their
“platforms”, marketplace banking subjects these services to competition on a non-discriminating
platform and thus potentially offers higher value to end customers. A conclusion is that the
business model of marketplace banking has disruptive qualities
A Sustainable Future with Autonomous Vehicles for Passenger Transport - The Means to Reach a Sustainable Future with Autonomous Vehicles through a Backcasting Approach
Today’s transportation system is associated with a number of negative impacts
when it comes to environmental, social, and economic sustainability. A prominent
negative impact is the emissions of greenhouse gases from vehicles, contributing to
climate change. With new technology like autonomous vehicles (AVs), a potential
to transform the way transportation is done today into something more sustainable
arises. Car manufacturers are paying more and more attention to AVs and it is likely
that this new technology will be diffused into society in the future. The aim of this
thesis is to lead the way and show how AVs can be used for passenger transportation
in a sustainable manner. The thesis scope is delimited to passenger transport only
and the geographical scope is limited to the city of Gothenburg.
The contribution of this thesis is to provide an alternative for how to get to a
sustainable transportation system in the city of Gothenburg in the future, with
the market introduction of AVs. This is done by using a backcasting approach,
where a sustainable future scenario, called our desired future, is identified in the
first step. Our desired future is based on a robot taxi service, where nobody owns
a car. The aim is to reduce resource consumption and to increase the utilization
of the existing capacity, by minimizing the number of vehicles and even out the
demand for transport. To provide a strategy for how to get to our desired future,
gaps between today and the desired future are identified and leverage points for how
to bridge the gaps are suggested. This is the first part of the thesis project.
In the second part of the project, the gap incentives to not own a car is chosen to
focus on. The chosen gap is seen as an important difference between the current
situation and our desired future, it is possible to start working with already today,
and it is crucial from a sustainability point of view, regardless of how long it will
take until fully autonomous vehicles are available on the market. Part 2 aims to find
a strategy for how to reduce the need for car ownership, to bridge the chosen gap,
which would get us one step closer to our desired future. By using the design thinking
methodology, a solution with the aim of decreasing the need for car ownership,
called the taxi platform, is developed. The taxi platform gathers the existing taxi
companies in an app and enables ride-sharing when beneficial. Hence, the taxi
platform is like our desired future, but with drivers instead of AVs. A strategy for
how to implement the suggested solution is provided in the form of a business model
canvas
Can vertical farms outgrow their cost? - An analysis of the competitive strength of vertical
Globalisation and urbanisation are creating sustainability challenges in the global food
system that require new methods for food production. Over 80 % of fruits and greens
consumed in Sweden are imported, accompanied with climate damaging side effects like
excess packaging, transportation, storing and cooling. With the introduction of vertical
farming (VF), urban agriculture is undergoing a transition towards drastically higher crop
yields and increased quality control in order to deliver more sustainable, resilient and safe
food production. VF could become a viable alternative to imported greens in Sweden,
benefitting both local producers and consumers.
The thesis aims to explore the financial, logistical and environmental prerequisites for
successfully implementing VF in Sweden, contributing to a more sustainable agricultural
industry. A literature review of business models, industry analysis, sustainability and
customer acceptance are used to create a theoretical framework. Data about vertical farming
actors is gathered through interviews using a qualitative research method, and two field
studies were conducted in Tokyo, Japan. The interview data was analysed with the help of
the theories and tools presented in the theoretical framework. Several environmental and
social benefits, concerning both the producer and other stakeholders, were identified and
point towards more sustainable agricultural practises. Overcoming the economic challenges
requires technological improvements to create entry barriers and mitigate bargaining
power, development in automatization, and reaching economies of scale. The competitive
edge over conventional farming methods comes from the ability to produce crops of
consistent quality and quantity year-round in close connection to the consumers.
The study indicates that Sweden is a promising country with the right prerequisites to adopt
vertical farming. The effect of implementing vertical farming will be a step forward in
reducing the environmental impact of agricultural practises while enabling a consistent
production of high-quality crops in urban environments
Investigation on alternative design of light-obscuring support structure for the secondary mirror
Imaging systems employing reflective elements for focusing light can be superior to traditional lens-based systems for use in broadband applications, or at wavelengths where lenses are unavailable. They are, however, hampered by stray intensity spikes caused by diffraction against the secondary mirror support structure. The use of curved vanes for reducing the intensity spikes present for conventional straight supports has previously been investigated[1, 2] but their implementation in short focal length systems, such as microscope objective lenses, has until now remained fairly unexplored. This thesis investigates the reduction of intensity spikes in 15X and 40X reflective Schwarzschild microscope objective lenses, where, among different support geometries, the use of curved vanes is further analysed.
By simulating diffraction of various support designs it was found that a constant curvature structure, comprising three arms, each with a subtended arc angle of 70° minimised the intensity spikes whilst still performing well with regard to other image quality criteria. In addition, the use of non-constant curvature structures for compensation of Gaussian beams was shown beneficial using numerical methods.
Practical implementation was carried out in the design of a 15X curved support structure Schwarzschild objective lens and was further experimentally tested and compared to a 15X Schwarzschild objective lens employing straight vanes. Results show a clear improvement in reduction of intensity spikes in favour of the curved vane structure. The measured data furthermore indicate a minor degradation with respect to other image quality criteria, in accordance with simulations