No description available.

No description available.

No description available.

Human Activity Recognition has become a popular research topic among data scientists. Over the years, multiple studies regarding humans and their daily motion habits have been investigated for many different purposes. This fact is not surprising when we look at all the opportunities and applications that can be applied and utilized thanks to the results of these algorithms. In this project we implement a system that can effectively collect sensor data from mobile devices, process it and by using supervised machine learning successfully predict the class of a performed activity. The project was executed based on datasets and features extracted from GPS sensors. The system was trained using various machine learning algorithms and Python SciKit to guarantee optimal solutions with accurate predictions. Finally, we applied a majority vote rule to secure the best possible accuracy of the activity classification process. As a result we were able to identify various activities including walking, cycling, driving and public transportation methods bus and metro with 90+% accuracy. / Att utföra aktivitetsigenkänning på människor har blivit ett populärt forskningsämne bland datavetare, där flertalet studier rörande människor och deras dagliga rörelsevanor undersökts för många olika syften. Detta är inte förvånande när man ser till de möjligheter och användningsområden som kan tillämpas och utnyttjas tack vare resultaten från dessa system. Detta projekt går ut på att implementera ett system som mha samlad sensordata från mobila enheter, kan bearbeta den och genom s.k övervakad maskininlärning med goda resultat bestämma den aktivitet som utförts. Projektet genomfördes baserat på dataset och egenskaper extraherade från GPS-data. Systemet tränades med olika maskininlärningsalgoritmer genom Python SciKit för att välja den bäst lämpade metoden för detta projekt. Slutligen tillämpade vi majority votemetoden för att säkerställa bästa möjliga noggrannhet i aktivitetsklassificeringsprocessen. Resultatet blev ett system som framgångsrikt kan identifiera aktiviteterna gå, cykla, köra bil samt med ett ytterligare fokus på kollektivtrafikmetoderna buss och tunnelbana, med en noggrannhet på över 90%. / Kandidatexjobb i elektroteknik 2020, KTH, Stockholm

This work proposes a new approach to fast detection of abnormal behaviour of cooling, IT, and power distribution systems in micro data centers based on machine learning techniques. Conventional protection of micro data centers focuses on monitoring individual parameters such as temperature at different locations and when these parameters reach certain high values, then an alarm will be triggered. This research employs machine learning techniques to extract normal and abnormal behaviour of the cooling and IT systems. Developed data acquisition system together with unsupervised learning methods quickly learns the physical dynamics of normal operation and can detect deviations from such behaviours. This provides an efficient way for not only producing health index for the micro data center, but also a rich label logging system that will be used for the supervised learning methods. The effectiveness of the proposed detection technique is evaluated on an micro data center placed at Computing Infrastructure Research Center (CIRC) in McMaster Innovation Park (MIP), McMaster University. / Thesis / Master of Science (MSc)

<p dir="ltr">Machine learning (ML) algorithms are increasingly being used in a variety of applications and are heavily relied upon to make decisions that impact people’s lives. ML models are often praised for their precision, yet they can discriminate against certain groups due to biased data. These biases, rooted in historical inequities, pose significant challenges in developing fair and unbiased models. Central to addressing this issue is the mitigation of biases inherent in the training data, as their presence can yield unfair and unjust outcomes when models are deployed in real-world scenarios. This study investigates the efficacy of data acquisition, i.e., one of the stages of data preparation, akin to the pre-processing bias mitigation technique. Through experimental evaluation, we showcase the effectiveness of data acquisition, where the data is acquired using data valuation techniques to enhance the fairness of machine learning models.</p>

The power grid has changed a great deal from what has been generally viewed as a traditional power grid. The modernization of the power grid has seen an increase in the integration and incorporation of computing and communication elements, creating an interdependence of both physical and cyber assets of the power grid. The fast-increasing connectivity has transformed the grid from what used to be primarily a physical system into a Cyber- Physical System (CPS). The physical elements within a power grid are well understood by power engineers; however, the newly deployed cyber aspects are new to most researchers and operators in this field. The new computing and communications structure brings new vulnerabilities along with all the benefits it provides. Cyber security of the power grid is critical due to the potential impact it can make on the community or society that relies on the critical infrastructure. These vulnerabilities have already been exploited in the attack on the Ukrainian power grid, a highly sophisticated, multi-layered attack which caused large power outages for numerous customers. There is an urgent need to understand the cyber aspects of the modernized power grid and take the necessary precautions such that the security of the CPS can be better achieved. The power grid is dependent on two main cyber infrastructures, i.e., Supervisory Control And Data Acquisition (SCADA) and Advanced Metering Infrastructure (AMI). This thesis investigates the AMI in power grids by developing a testbed environment that can be created and used to better understand and develop security strategies to remove the vulnerabilities that exist within it. The testbed is to be used to conduct and implement security strategies, i.e., an Intrusion Detections Systems (IDS), creating an emulated environment to best resemble the environment of the AMI system. A DoS flooding attack and an IDS are implemented on the emulated testbed to show the effectiveness and validate the performance of the emulated testbed. / M.S. / The power grid is becoming more digitized and is utilizing information and communication technologies more, hence the smart grid. New systems are developed and utilized in the modernized power grid that directly relies on new communication networks. The power grid is becoming more efficient and more effective due to these developments, however, there are some considerations to be made as for the security of the power grid. An important expectation of the power grid is the reliability of power delivery to its customers. New information and communication technology integration brings rise to new cyber vulnerabilities that can inhibit the functionality of the power grid. A coordinated cyber-attack was conducted against the Ukrainian power grid in 2015 that targeted the cyber vulnerabilities of the system. The attackers made sure that the grid operators were unable to observe their system being attacked via Denial of Service attacks. Smart meters are the digitized equivalent of a traditional energy meter, it wirelessly communicates with the grid operators. An increase in deployment of these smart meters makes it such that we are more dependent on them and hence creating a new vulnerability for an attack. The smart meter integration into the power grid needs to be studied and carefully considered for the prevention of attacks. A testbed is created using devices that emulate the smart meters and a network is established between the devices. The network was attacked with a Denial of Service attack to validate the testbed performance, and an Intrusion detection method was developed and applied onto the testbed to prove that the testbed created can be used to study and develop methods to cover the vulnerabilities present.

The development of an add-on sensorless position estimator for a 4.8 m Linear Switched Reluctance Machine (LSRM) with minimal modifications to the transducer-based controller is investigated for the first time in this study. LSRMs require position feedback for closed-loop control but present a low cost, high energy efficiency alternative for linear actuation due to their rugged construction and single-sided excitation. Mechanical position transducers mounted on the vehicle are expensive and can impact reliability. The use of a sensorless position estimator removes all electronics from the passive vehicle, resulting in considerable reductions in cost, maintenance, and mechanical complexity. This study examines the use of an add-on processor and data acquisition system for sensorless position estimation. An approach exploiting the active phase windings is used to preserve the normal operation of the transducer-based DSP controller with the goal of limiting reductions in high performance features such as force ripple reduction and velocity control [3]. The estimator system is retrofit to the transducer-based DSP controller by mimicking the output of a mechanical position sensor by emulating a Quadrature encoder. The feasibility and design issues for an add-on or retrofit position estimator are investigated. Although sensorless schemes for rotary Switched Reluctance Machines (SRMs) have been studied in detail, the problem of sensorless implementations for LSRMs has not been addressed. Experimental validation of the proposed sensorless estimation scheme is attempted, but closed-loop operation is not achieved successfully due to air gap fluctuations. In depth analysis of the sources and propagation of error is presented. / Master of Science

To prevent further brain tumour growth, malignant tissue should be removed as completely as possible in neurosurgical operations. Therefore, differentiation between tumour and brain tissue as well as detecting functional areas is very important. Hyperspectral imaging (HSI) can be used to get spatial information about brain tissue types and characteristics in a quasi-continuous reflection spectrum. In this paper, workflow and some aspects of an adapted hardware system for intraoperative hyperspectral data acquisition in neurosurgery are discussed. By comparing an intraoperative with a laboratory setup, the influences of the surgical microscope are made visible through the differences in illumination and a pixel- and wavelength-specific signal-to-noise ratio (SNR) calculation. Due to the significant differences in shape and wavelength-dependent intensity of light sources, it can be shown which kind of illumination is most suitable for the setups. Spectra between 550 and 1,000 nm are characterized of at least 40 dB SNR in laboratory and 25 dB in intraoperative setup in an area of the image relevant for evaluation. A first validation of the intraoperative hyperspectral imaging hardware setup shows that all system parts and intraoperatively recorded data can be evaluated. Exemplarily, a classification map was generated that allows visualization of measured properties of raw data. The results reveal that it is possible and beneficial to use HSI for wavelength-related intraoperative data acquisition in neurosurgery. There are still technical facts to optimize for raw data detection prior to adapting image processing algorithms to specify tissue quality and function.:Abstract Introduction Materials and methods (Clinical workflow and setup for hyperspectral imaging process, Characteristics of the lighting, Characteristics of the hyperspectral imaging camera, Spectral data acquisition and raw data pre-processing in neurosurgery, Spectral data evaluation) Results (Spectral characteristics of the lighting, SNR of the HSI camera, Data acquisition and raw data preprocessing during neurosurgical operation, Spectral data evaluation) Discussion Conclusions

Актуальность темы дипломной работы «Разработка центральной управляющей системы» обусловлена необходимостью повышения эффективности работы эксплуатационных и сервисных служб. Внедрение этой системы позволит значительно улучшить качество обслуживания клиентов и оптимизировать процессы управления техническими системами зданий. Система будет способна автономно решать возникающие инциденты или ассистировать персоналу в их устранении, обеспечивая надежную и бесперебойную работу всех сервисов. / The relevance of the thesis topic "Development of a Central Control System" is driven by the need to enhance the efficiency of operational and maintenance services. Implementing this system will significantly improve the quality of customer service and optimize the management processes of building technical systems. The system will be capable of autonomously resolving incidents or assisting personnel in their resolution, ensuring reliable and uninterrupted operation of all services.