Abstract

Abstract
Museums promote cultural experiences through the exhibits and the stories behind them. Nevertheless, museums are not always designed to engage and interest young audiences, particularly teenagers. This Ph.D. proposal in Digital Media explores how digital technologies can facilitate Natural History and Science Museums in fostering and creating immersive museum experiences for teenagers. Especially by using digital storytelling along with location-based gaming. The overall objectives of the work are to establish guidelines, design, develop and study interactive storytelling and gamification experiences in those type of museums focusing in particular on delivering pleasurable and engaging experiences for teens of 15-17 years old.

Abstract
This paper presents a novel multi-stage stochastic distributed generation investment planning model for making investment decisions under uncertainty. The problem, formulated from a coordinated system planning viewpoint, simultaneously minimizes the net present value of costs rated to losses, emission, operation, and maintenance, as well as the cost of unserved energy. The formulation is anchored on a two-period planning horizon, each having multiple stages. The first period is a short-term horizon in which robust decisions are pursued in the face of uncertainty; whereas, the second one spans over a medium to long-term horizon involving exploratory and/or flexible investment decisions. The operational variability and uncertainty introduced by intermittent generation sources, electricity demand, emission prices, demand growth, and others are accounted for via probabilistic and stochastic methods, respectively. Metrics such as cost of ignoring uncertainty and value of perfect information are used to clearly demonstrate the benefits of the proposed stochastic model. A real-life distribution network system is used as a case study and the results show the effectiveness of the proposed model.

Abstract
In stress sensing, Window-derived Heart Rate Variability (W-HRV) methods are by far the most heavily used feature extraction methods. However, these W-HRV methods come with a variety of tradeoffs that motivate the development of alternative methods in stress sensing. We compare our method of using HeartBeat Morphology (HBM) features for stress sensing to the traditional W-HRV method for feature extraction. In order to adequately evaluate these methods we conduct a Trier Social Stress Test (TSST) to elicit stress in a group of 13 firefighters while recording their ECG, actigraphy, and psychological self-assessment measures. We utilize the data from this experiment to analyze both feature extraction methods in terms of computational complexity, detection resolution performance, and event localization performance. Our results show that each method has an ideal niche for its use in stress sensing. HBM features tend to be more effective in an online, stress detection context. W-HRV shows to be more suitable for offline post processing to determine the exact localization of the stress event.

Abstract
The fast economic growth of China along the last two decades has created a strong impact on the environment. The occurrence of heavy haze pollution days is the most visible effect. Although many researchers have studied such problem, a high number of spatio-temporal limitations in the recent studies were identified. From our best knowledge the long trends of PM2.5 concentrations were not fully investigated in China, in particular the year-to-year trends and the seasonal and daily cycles. Therefore, in this work the PM2.5 concentrations collected from automatic monitors from five urban sites located in megacities with different climatic zones in China were analysed: Beijing (40 degrees N), Chengdu (31 degrees N), Guangzhou (23 degrees N), Shanghai (31 degrees N) and Shenyang (43 degrees N). For an inter-comparison a meta-analysis was carried out. An evaluation conducted since 1999 demonstrates that PM2.5 concentrations have been reduced until 2008, period which match with the occurrence of the Olympic Games. However, a seasonal analysis highlight that such decrease occurs mostly during warmer seasons than cold seasons. During winter PM2.5 concentrations are typically 1.3 to 2.7 higher than in summer. The average daily cycle shows that the lowest and highest PM2.5 concentrations often occurs in the afternoon and evening hours respectively. Such daily variations are mostly driven by the daily variation of the boundary layer depth and emissions. Although the PM2.5 levels have showing signs of improvement, even during the warming season the values are still too high in comparison with the annual environmental standards of China (35 mu g m(-3)). Moreover, during cold seasons the north regions have values twice higher than this limit. Thus, to fulfil these standards the governmental mitigation measures need to be strongly reinforced in order to optimize the daily living energy consumption, primarily in the north regions of China and during the winter periods.

Abstract
This paper describes a study on applying data mining techniques to power transformer failure prediction. The data set used consisted not only on DGA tests, but also in other tests done to the transformer’s insulating oil. This dataset presented several challenges, such as highly imbalanced classes (common in failure prediction problems), and the temporal nature of the observations. To overcome these challenges, several techniques were applied for prediction and better understand the dataset. Pre-processing and temporality incorporation in the dataset is discussed. For prediction, a 1-class and 2-class SVM, decision trees and random forests, as well as a LSTM neural network were applied to the dataset. As the prediction performance was low (high false-positive rate), we conducted a test to ascertain if the amount of data collected was sufficient. Results indicate that the frequency of data collection was not adequate, hinting that the degradation period was shorter than the periodicity of data collection.