Results 1 - 10 of 57
Results 1 - 10 of 57. Search took: 0.021 seconds
|Sort by: date | relevance|
[en] Highlights: • An autoencoder-based ensemble method is developed for anomaly detection. • Autoencoders can capture the intrinsic characteristics in building energy data. • The performance of various autoencoder types and training schemes is compared. • Methods are developed for performance evaluation without using anomaly labels. • This study provides data-driven solutions to unsupervised anomaly detection. - Abstract: Practical building operations usually deviate from the designed building operational performance due to the wide existence of operating faults and improper control strategies. Great energy saving potential can be realized if inefficient or faulty operations are detected and amended in time. The vast amounts of building operational data collected by the Building Automation System have made it feasible to develop data-driven approaches to anomaly detection. Compared with supervised analytics, unsupervised anomaly detection is more practical in analyzing real-world building operational data, as anomaly labels are typically not available. Autoencoder is a very powerful method for the unsupervised learning of high-level data representations. Recent development in deep learning has endowed autoencoders with even greater capability in analyzing complex, high-dimensional and large-scale data. This study investigates the potential of autoencoders in detecting anomalies in building energy data. An autoencoder-based ensemble method is proposed while providing a comprehensive comparison on different autoencoder types and training schemes. Considering the unique learning mechanism of autoencoders, specific methods have been designed to evaluate the autoencoder performance. The research results can be used as foundation for building professionals to develop advanced tools for anomaly detection and performance benchmarking.
[en] Highlights: • Smart home energy management in summer season is considered. • Thermal and electrical performance of smart home is studied under price uncertainty. • Market price uncertainty is modeled based on information gap decision theory. • Smart appliances, battery, water tank, backup boiler and fuel cell are the equipment. - Abstract: After restructuring in the electricity industry, some concepts such as smart home and renewable energy sources improved extensively. This paper proposes information gap decision theory (IGDT) technique for robust energy management of smart home in summer season in the presence of market price fluctuation. Therefore, the proposed model is practical in a realistic model. The IGDT method contains the robustness and opportunity functions. The harmful aspect of price uncertainty is modeled by robustness function and the beneficial aspect of price uncertainty is modeled by opportunity function. The proposed IGDT-based performance optimization problem of low-energy smart home is formulated as mixed-integer non-linear programming (MINLP) and solved by General Algebraic Modeling System (GAMS) optimization software. Two scenarios as normal and smart scenarios are used to investigate the proposed model.
[en] Highlight• Analysis of the UK energy and non-energy intensive sectors firms in a real business cycle (RBC) model. We investigate the role of energy shocks during the Great Recession. We study the behaviour of the UK energy and non-energy intensive sectors firms in a real business cycle (RBC) model using unfiltered data. The model is econometrically estimated and tested by indirect inference. Output contraction during the Great Recession was largely caused by energy price and sector-specific productivity shocks, all of which are non-stationary and hence tend to dominate the sample variance decomposition. We also found that the channel by which the energy price shock reduces output in the model is via the terms of trade: these fall permanently when world energy prices increase and as substitutes for energy inputs are strictly limited there are few reactions via production channels. Therefore, there is no other way to balance the deteriorating current account than through lower domestic absorption.
[en] The nuclear education problem is treated as societal optimization task of nuclear energy management, with the key parameter of optimization—stakeholder awareness level. As the key principles of optimisation are chosen: a self-organization concept, the principle of the requisite variety, where as a primary source of growth of internal variety is information and knowledge. We have shown: public education, social learning and the use of mass media are efficient self-organization mechanisms, thereby forming a knowledge-creating community. Such a created knowledge could facilitate solution of key issues: a) public acceptance of novel nuclear objects, b) promotion of adequate risk perception, and c) fostering of interest to nuclear energy. Comprehensive knowledge management and informational support firstly is needed in: a) for increasing general nuclear awareness and confidence level to nuclear activities, b) personnel education and training, c) reliable staff renascence, d) public education and involvement of all stakeholder categories in decision making, e) risk management. A common approach to nuclear education should include also comprehensive research activities, thereby joining knowledge acquisition with the generation of novel advanced knowledge. (author)
[en] The Nuclear Energy Management School is proposed as a good tool to structure the experiences of industries. The importance of a short-term international education programme for gathering knowledge regarding nuclear embarkation projects is discussed in this paper. The results of evaluating education efficiency from 2013 to 2016 will also be introduced in this presentation (or poster). (author)
BackgroundSmart grid tools (e.g., individualized disaggregated data, goal setting, and behavioural suggestions/feedback) increase opportunities to reduce or shift residential electricity consumption, but can they shape residential energy culture? And what underlying factors influence this shift? Insights are identified from a qualitative analysis of a 3-year residential smart grid project in a suburb of Toronto, Canada. Interviews evaluated whether participants experienced changes in their energy culture and identified underlying factors. In particular, the impacts of the project tools on participants’ norms (attitudes and awareness towards energy management), material culture (technical changes) and energy practices (conservation/peak shifting actions) were assessed, and motivations and barriers towards energy management were identified. The effectiveness of engagement mechanisms (i.e., web portal, reminder emails, webinars, incentivized control programme, and weekly electricity reports) was also evaluated. By examining detailed qualitative feedback following a multi-year suburban smart grid project in Ontario this study aims to (1) assess the changes in energy culture over the duration of the 3-year project and to (2) assess the underlying factors influencing household energy consumption and a smart residential energy culture.
ResultsFindings from the interview were compared to the results of an initial project survey to identify longer-term influences on energy culture. Increases in self-reported awareness and practices were accounted for, with the web portal and individualized weekly feedback email reported most frequently as causes of change. While increased awareness was obtained, participants needed additional guidance to make substantial changes. Although participants were financially motivated, norms of lifestyle and convenience, as well as competing household values of energy management were the largest barriers to home energy management.
ConclusionsThis study showcases challenges for engaging homeowners with home energy management technologies due to norms as well as competing household interests. Nuanced findings as an outcome of this study framed around energy cultures can influence future studies on smart grid engagement and consumer behaviour with larger samples sizes. In particular, future studies can further investigate the motivations and barriers surrounding residential energy cultures, how to engage different ‘cultures of consumption’ within households, and elements to effectively educate consumers beyond disaggregated feedback.
[en] Close mutual cooperation among nuclear-related organizations, such as government, industry and academia is extremely useful to promote Nuclear Human Resources Development (HRD). National HRD network has already been established in Japan since Nov. 20101. The network has promoted the following five discussion on (1) Elementary to High School Education (2) Nuclear Education at Universities and Colleges, (3) HRD for Working Engineers, (4) HRD to Internationalize National Human Resources, and (5) Supportive HRD Activities to newly NPP Introducing Countries successfully. Through the establishment of the network, the communication has been strongly improved so that the Japan-IAEA joint Nuclear Energy Management School can be held successfully every year. Based on the good experience with the network, Japan would like to recommend the introduction of national Nuclear HRD (N-HRD)-Network to the NPP-embarking Countries. We are interested in cooperation with IAEA for establishment of National N-HRD Network for efficient and effective N-HRD. (author)
[en] Highlights: • A bi-level model for a microgrid power and reserve capacity planning is developed. • The model is cast within the context of a distribution system operator (DSO). • The DSO and microgrid relationship is established in a structural/economical manner. • Results obtained show bi-level optimization decreases overall operating cost. - Abstract: This paper proposes a bi–level formulation for a coupled microgrid power and reserve capacity planning problem, cast within the jurisdiction of a distribution system operator(DSO). The upper level problem of the proposed bi–level model represents a microgrid planner whose goal is to minimize its planning and operational cost, while the lower level problem represents a DSO whose primary duty is to ensure reliable power supply. The microgrid planner, pursues its interest by co–optimizing the design configuration and power output of individual distributed energy resources (DERs), while the DSO maximizes the capacity of flexible reserve resources. The proposed model is recast as a mathematical program with equilibrium constraints (MPEC) wherein the decision variables of the two problems are independently controlled. Application of the proposed approach to the energy infrastructure of a Canadian utility network is discussed. Results obtained through its application are compared to an alternative multi–objective planning model and the improved benefits are assigned to the corresponding stakeholders.
[en] Highlights: • Fuel economy for Fuel Cell Hybrid Power Systems using optimal and sub-optimal strategies. • Fuel economy using Real-Time Optimization strategies with current slope limiter. • 6 kW fuel cell under static feed-forward and load-following control is the reference. • Fuel economy is obtained in the entire range of load demand for proposed strategies. • Up to 5 L fuel economy for optimal strategy compared to sub-optimal ones. - Abstract: The aim of this paper is to compare an optimal and a sub-optimal strategy for the Fuel Cell Hybrid Power Systems based on Maximum Power Point tracking algorithms (with global feature or not) with the basic energy management strategy, namely the static Feed-Forward strategy considered as reference. The fuel economy is used as the unique performance indicator. The gaps in fuel economy for two Real-Time Optimization strategies based on Global Extremum Seeking algorithm and Perturb & Observe algorithm are compared to highlight the advantages of the global optimization strategies. Up to 5 L fuel economy was obtained for optimal strategies compared to sub-optimal ones. Also, the gaps in fuel economy are estimated for the proposed strategies using two levels of the FC current slope. The results of this study obtained for constant load are validated on a variable and unknown profile of the load power as well.