Building energy models are used to forecast building energy use to design and control efficient building energy systems. Building energy use can generally be decomposed into heating, ventilation and air conditioning, refrigeration, appliance and lighting loads. These loads will depend on multiple factors such as outdoor weather conditions, occupants, building type, controls and scheduling. Data-driven models have become more popular with the increase in smart meter data available that can be used to train and fit the models. Additionally, buildings with high refrigeration loads have greater heat harvesting potential, however, few data-driven models have been developed for buildings such as supermarkets and ice rinks. In this work, linear regression models are used to predict the disaggregated space cooling, heating, baseload and refrigeration components of building energy use. In most cases, measured aggregate electricity use is available, however individual appliances or component loads require submetering equipment which can be expensive. Therefore the proposed models use time-based and weather features to separate the thermal and baseload portions of the electrical load. A generalized approach is also used to predict new buildings with data from existing buildings. Furthermore, a simplified model is used to predict hourly space heating from monthly natural gas measurements and hourly weather measurements. The models were evaluated on real data from buildings in Ontario and the disaggregated loads were verified with synthetic data. The results found that aggregate use was predicted reasonably well using linear regression methods, with most building types having a median normalized root mean squared error between 0.2 and 0.3, depending on the forecasting period. The model is flexible as it does not require detailed information related to the building such as lighting or setpoint schedules, however, it can be adapted in the future to include additional information and improve predictive capability. / Thesis / Master of Applied Science (MASc)

A variety of energy models and tools have been used for an comprehensive analysis of the complex energy systems and the design of pathway to sustainable energy world. This thesis analyzes three interesting problems in the electricity sector by developing and using suitable energy models. Chapter 2 investigates how to incorporate demand responsiveness for policy analysis in the electricity sector using a least-cost model. This study develops its own least-cost model which includes some characteristics for two important policies in the electricity sector, and suggests an iterative approach for incorporating the demand response to price change under new policy. Based on a case study, the state of Georgia, this chapter shows the effects of including demand response on the evaluation of policy. Chapter 3 is about new technology adoption pathways in the electric power system. In this chapter, by investigating the related status of policies and specifications of electric vehicles and wind power technologies in the U.S., several adoption pathways of the technologies in the U.S. eastern interconnection have been developed. This study develops four-serial models for the estimation of future economic and environmental impacts of the technologies' penetration. The results show that the total greenhouse gas emissions of the entire energy system do not substantially decrease even with a high level of electric vehicle adoption. The combination of two technologies, even more with appropriate policies, can notably decrease the total greenhouse gas emissions. Chapter 4 is a study about demand response programs, particularly optional time-based rates, for residential customers. This chapter analyzes the main reason that the participation of the current programs is low even though the programs have benefits. This study investigates two policy tools, a subsidy for flexible residential demand and a shared-savings mechanism based on consumption pattern changes, and examines the implementation of the tools and their potential to overcome the current inefficient operation.

Orientador: Sergio Valdir Bajay / Tese (doutorado) - Universidade Estadual de Campinas, Faculdade de Engenharia Mecanica / Made available in DSpace on 2018-08-06T05:07:08Z (GMT). No. of bitstreams: 1 Carvalho_ClaudioBezerrade_D.pdf: 1717945 bytes, checksum: 278bbb29104ec96389a1e6616a1dc209 (MD5) Previous issue date: 2005 / Resumo: Este trabalho traz uma avaliação critica do planejamento energético de longo prazo realizado no Pais nos últimos anos e aponta tanto para a necessidade de uma melhor integração entre as atividades de planejamento energético, elaboração de políticas públicas e regulação dos mercados de energia, como para uma premente descentralização do processo de planejamento. Como resultados desta análise e com base em experiências bem sucedidas no exterior, são propostos avanços metodológicos para a elaboração de futuras projeções e o desenvolvimento de um modelo integrado de projeção da demanda e da oferta de energéticos. Como a aplicação de tal modelo está vinculada à utilização de uma base de dados ampla e consistente, é proposto o desenvolvimento de um sistema nacional de informações energéticas, integrado a um sistema de informações executivas, cujo objetivo é servir de suporte para as atividades desenvolvidas pelo Ministério de Minas e Energia. Discute-se os vários métodos de tratamento das incertezas nos modelos energéticos, com destaques para a elaboração de cenários alternativos de desenvolvimento e para o uso da técnica Delphi de levantamento de opiniões de especialistas. Monta-se, por fim, à guisa de um estudo de caso que visa contribuir para o necessário processo de descentralização do planejamento energético no País, cenários alternativos de desenvolvimento para a projeção da demanda energética do Estado da Bahia, de uma forma concatenada com cenários semelhantes no âmbito nacional / Abstract: This work brings a critical evaluation of the long-term energy planning carried out in the country in the last years, pointing out both for the need of a better integration of the activities concerning energy planning, policy making and regulation of energy markets, and for an urgent decentralization of the planning process. As results of this analysis and based on successful experiences abroad,methodological advances are proposed for the elaboration of future forecasts, together with the development of an integrated model for forecasting energy demand and supply. As the application of such a model requires a broad and consistent data basis, setting up a national system of energy information is proposed, integrated to a system of executive information, aimed to support the activities of the Ministry of Mines and Energy. The several methods for treating uncertainties in energy modeling are discussed, with emphasis on the elaboration of alternative development scenarios and the use of the Delphi technique for collecting and processing the opinions of specialists. At the end, alternative development scenarios for forecasting the energy demand in the State of Bahia, linked to similar scenarios at the national level, are elaborated, as a study case aimed to contribute for the necessary decentralization process of energy planning in the country / Doutorado / Planejamento de Sistemas Energeticos / Doutor em Planejamento de Sistemas Energéticos

Cette recherche se base sur l’établissement de différents critères de flexibilité professionnelle lors du changement de prescription en 2011 par les enseignants de Sciences Industrielles de l’Ingénieur et sur la comparaison des méthodes de modélisation de l’énergie entre la Technologie et les Sciences Physiques. La méthodologie utilisée dans la thèse est basée sur l’analyse de cinq types de données d’analyse : le recueil des difficultés à enseigner des savoirs spécifiques, le recueil des difficultés spécifiques à l’enseignement de l’ETT et les raisons de ces difficultés, des enregistrements des interactions entre les enseignants et les modélisations des enseignants lors d’une simulation de séance, le recueil des difficultés de compréhension des savoirs relatifs à l’énergie des élèves et le recueil des difficultés potentielles en ETT des élèves. Les résultats montrent que les enseignants ne déclarent pas les mêmes niveaux de difficultés pour enseigner l’énergie en fonction de leur spécialité. Les élèves de STI2D déclarent des niveaux de difficultés variables vis-à-vis de l’apprentissage de l’énergie en ETT / This research focuses on the setting of different flexibility’s criteria established by engineering science teachers (task’s acceptation, motivation, utility perceived …) and on the comparison of methods of modeling energy between Technology and Physics. The methodology used is based on the analysis of five types of data: the collection of difficulties to teach some knowledge, the collection of difficulties specific to energy teaching, some records of interactions between teachers during a classroom simulation, the collection of students' difficulties in terms of energy concepts and the collection of potential difficulties in technological education. The results show that the teachers don’t report the same levels of difficulty for teaching energy according to their specialty as the students who report varying levels of difficulty to understand energy

This thesis investigates the integration of Stein Variational Gradient Descent (SVGD) with Joint Energy Models (JEMs), comparing its performance to Stochastic Gradient Langevin Dynamics (SGLD). We incorporated a generative loss term with an entropy component to enhance diversity and a smoothing factor to mitigate numerical instability issues commonly associated with the energy function in energy-based models. Experiments on the CIFAR-10 dataset demonstrate that SGLD, particularly with Sharpness-Aware Minimization (SAM), outperforms SVGD in classification accuracy. However, SVGD without SAM, despite its lower classification accuracy, exhibits lower calibration error underscoring its potential for developing well-calibrated classifiers required in safety-critical applications. Our results emphasize the importance of adaptive tuning of the SVGD smoothing factor ($alpha$) to balance generative and classification objectives. This thesis highlights the trade-offs between computational cost and performance, with SVGD demanding significant resources. Our findings stress the need for adaptive scaling and robust optimization techniques to enhance the stability and efficacy of JEMs. This thesis lays the groundwork for exploring more efficient and robust sampling techniques within the JEM framework, offering insights into the integration of SVGD with JEMs. / Master of Science / This thesis explores advanced techniques for improving machine learning models with a focus on developing well-calibrated and robust classifiers. We concentrated on two methods, Stein Variational Gradient Descent (SVGD) and Stochastic Gradient Langevin Dynamics (SGLD), to evaluate their effectiveness in enhancing classification accuracy and reliability. Our research introduced a new mathematical approach to improve the stability and performance of Joint Energy Models (JEMs). By leveraging the generative capabilities of SVGD, the model is guided to learn better data representations, which are crucial for robust classification. Using the CIFAR-10 image dataset, we confirmed prior research indicating that SGLD, particularly when combined with an optimization method called Sharpness-Aware Minimization (SAM), delivered the best results in terms of accuracy and stability. Notably, SVGD without SAM, despite yielding slightly lower classification accuracy, exhibited significantly lower calibration error, making it particularly valuable for safety-critical applications. However, SVGD required careful tuning of hyperparameters and substantial computational resources. This study lays the groundwork for future efforts to enhance the efficiency and reliability of these advanced sampling techniques, with the overarching goal of improving classifier calibration and robustness with JEMs.

The United Nations’ Sustainable Development Goal 7 (SDG7) of Agenda 2030 calls for an increase in the use of renewable energy sources, among other targets. The percentage of fossil fuel-fired thermal generation for electricity is increasingly being reduced as renewable energy technologies (RET) advance in cost-competitiveness, and as greenhouse gas and industrial air pollutant emission limits become more stringent. In certain cases, renewable energy contributes to energy security by improving a nation’s trade balance, since local resources are harnessed and imports are reduced. RET investments are becoming more frequent gaining a sizeable share in the electric power mix of numerous countries. However, RET is affected by existing fossil fuel-fired electricity generation, especially in countries that have domestic reserves. While coal may be dirty, others such as natural gas provide multiple benefits, presenting a challenge to renewables. Additionally, RET endowment varies for each geographical location. This often does not correspond to the location of major electricity demand centers.  Therefore, large scale RET adoption and integration becomes logistically more cumbersome, as it necessitates existence of a developed grid network. Utilizing a series of analyses in two different settings – Africa and Cyprus – this thesis draws insights on RET growth policy and the level of technology representation in long term energy models. In order to capture specific challenges of RET integration, enhancements in traditional long-term energy system models are called for and carried out.  The case of Africa is used to assess adoption of RET under various trade scenarios. It is home to some of the world’s greatest RET resource potential and the single largest potential RET project, Grand Inga.  While, the island of Cyprus has goals of introducing large percentages of RET into its electric power mix. Each have important idiosyncrasies which are reflected in the analysis. On the one hand, natural gas competes with RET in Cyprus and forms a key transition fuel away from oil. On the other hand, lack of cross-border interconnectors limit RET project development across Africa. / <p>QC 20170519</p>

Energy management has been an important function of organisations since the oil crisis of the mid 1970’s led to hugely increased costs of energy. Although the financial costs of energy are still important, the growing recognition of the environmental costs of fossil-fuel energy is becoming more important. Legislation is also a key driver. The UK has set an ambitious greenhouse gas (GHG) reduction target of 80% of 1990 levels by 2050 in response to a strong international commitment to reduce GHG emissions globally. This work is concerned with the management of energy consumption in buildings through the analysis of energy consumption data. Buildings are a key source of emissions with a wide range of energy-consuming equipment, such as photocopiers or refrigerators, boilers, air-conditioning plant and lighting, delivering services to the building occupants. Energy wastage can be identified through an understanding of consumption patterns and in particular, of changes in these patterns over time. Changes in consumption patterns may have any number of causes; a fault in heating controls; a boiler or lighting replacement scheme; or a change in working practice entirely unrelated to energy management. Standard data analysis techniques such as degree-day modelling and CUSUM provide a means to measure and monitor consumption patterns. These techniques were designed for use with monthly billing data. Modern energy metering systems automatically generate data at half-hourly or better resolution. Standard techniques are not designed to capture the detailed information contained in this comparatively high-resolution data. The introduction of automated metering also introduces the need for automated analysis. This work assumes that consumption patterns are generally consistent in the short-term but will inevitably change. A novel statistical method is developed which builds automated event detection into a novel consumption modelling algorithm. Understanding these changes to consumption patterns is critical to energy management. Leicester City Council has provided half-hourly data from over 300 buildings covering up to seven years of consumption (a total of nearly 50 million meter readings). Automatic event detection pinpoints and quantifies over 5,000 statistically significant events in the Leicester dataset. It is shown that the total impact of these events is a decrease in overall consumption. Viewing consumption patterns in this way allows for a new, event-oriented approach to energy management where large datasets are automatically and rapidly analysed to produce summary meta-data describing their salient features. These event-oriented meta-data can be used to navigate the raw data event by event and are highly complementary to strategic energy management.

No description available.