Abstract

Abstract
In conventional hybrid energy storage systems, two storage units complement each other. One low-capacity and fast-response unit as a power supplier, and one high-capacity and low-response unit as an energy supplier. The power supplier mitigates fast fluctuations in generation or demand by transferring energy over seconds or minutes, and the energy supplier transfers energy over hours for managing energy. According to this concept, this paper presents a new model of hybrid energy storage systems, where three energy suppliers are considered as a three-level hybrid energy storage system. Energy storage at level 1 shifts energy from off-peak (or low-cost) hours to the on-peak (or high-cost) hours during one day, the storage unit at level 2 transfers energy from off-peak (or low-cost) days to the on-peak (or high-cost) days for the period of one week, and level 3 transfers energy from off-peak seasons to the on-peak seasons through one year. The proposed planning results in a large-scale optimization programming that optimizes large numbers of design variables at the same time. In order to increase the flexibility of the planning, the initial energy of the storage units is also modeled as a design variable and optimized. The uncertainty of loads is modeled and a stochastic planning is carried out to solve the problem. The introduced three-level hybrid energy storage planning is simulated on two test systems, and the results demonstrate that the proposed planning can reduce the planning cost by about 1.8%.

Abstract
This paper addresses the network expansion planning of an active microgrid that utilizes Distributed Energy Resources (DERs). The microgrid uses Combined Cooling, Heating and Power (CCHP) systems with their heating and cooling network. The proposed method uses a bi-level iterative optimization algorithm for optimal expansion and operational planning of the microgrid that consists of different zones, and each zone can transact electricity with the upward utility. The transaction of electricity with the upward utility can be performed based on demand response programs that consist of the time-of-use program and/or direct load control. DERs are CHPs, small wind turbines, photovoltaic systems, electric and cooling storage, gas fired boilers and absorption and compression chillers are used to supply different zones' electrical, heating, and cooling loads. The proposed model minimizes the system's investment, operation, interruption and environmental costs; meanwhile, it maximizes electricity export revenues and the reliability of the system. The proposed method is applied to a real building complex and five different scenarios are considered to evaluate the impact of different energy supply configurations and operational paradigm on the investment and operational costs. The effectiveness of the introduced algorithm has been assessed. The implementation of the proposed algorithm reduces the aggregated investment and operational costs of the test system in about 54.7% with respect to the custom expansion planning method.

Abstract
During recent decades with the power system restructuring process, centralized energy sources are being replaced with decentralized ones. This phenomenon has resulted in a novel concept in electric power systems, particularly in distribution systems, known as Distributed Generation (DG). On one hand, utilizing DG is important for secure power generation and reducing power losses. On the other hand, widespread use of such technologies introduces new challenges to power systems such as their optimal location, protection devices' settings, voltage regulation, and Power Quality (PQ) issues. Another key point which needs to be considered relates to specific DG technologies based on Renewable Energy Sources (RESs), such as wind and solar, due to their uncertain power generation. In this regard, this paper provides a comprehensive review of different types of DG and investigates the newly emerging challenges arising in the presence of DG in electrical grids.

Abstract
This paper proposes a stochastic decision making problem for a wind power producer (WPP) in the day-ahead (DA) and balancing markets. In this problem, bidding strategy of the WPP in a competitive electricity market and also its participation to supply demand response (DR) and electric vehicle (EV) aggregators is determined to achieve the maximum profit. In this model, DR and EV aggregators are able to choose the most competitive WPP in such a way that their energy payments be minimized in the scheduling horizon. Therefore, the problem is formulated as a stochastic bi-level programming model with conflict objectives of the WPP and the aggregators. Moreover, owing to the uncertainties associated with market prices, offered prices by rival WPPs, demand of DR and EV aggregators, conditional value at risk (CVaR) is applied to the proposed model. The attained stochastic bi-level problem is transformed to a linear stochastic single level problem with equilibrium constraints using Karush Kuhn Tucker (KKT) optimality conditions. The proposed model is evaluated on a realistic case study and the impacts of risk-averse behavior and demand response participants on the decision making problem of the WPP are investigated. Numerical results indicate that with increasing DR participants of 0%, 60% and 100%, CVaR of WPP increases 33.81%, 40.79% and 46.99%, respectively. This means that if the loads are more responsive, the WPP tries to control the profit variability due to the uncertainties of loads.

Abstract
Accurate solar photovoltaic power forecasting can help mitigate the potential risk caused by the uncertainty of photovoltaic out power in systems with high penetration levels of solar photovoltaic generation. Weather classification based photovoltaic power forecasting modeling is an effective method to enhance its forecasting precision because photovoltaic output power strongly depends on the specific weather statuses in a given time period. However, the most intractable problems in weather classification models are the insufficiency of training dataset (especially for the extreme weather types) and the selection of applied classifiers. Given the above considerations, a generative adversarial networks and convolutional neural networks-based weather classification model is proposed in this paper. First, 33 meteorological weather types are reclassified into 10 weather types by putting several single weather types together to constitute a new weather type. Then a data-driven generative model named generative adversarial networks is employed to augment the training dataset for each weather types. Finally, the convolutional neural networks-based weather classification model was trained by the augmented dataset that consists of both original and generated solar irradiance data. In the case study, we evaluated the quality of generative adversarial networks-generated data, compared the performance of convolutional neural networks classification models with traditional machine learning classification models such as support vector machine, multilayer perceptron, and k-nearest neighbors algorithm, investigated the precision improvement of different classification models achieved by generative adversarial networks, and applied the weather classification models in solar irradiance forecasting. The simulation results illustrate that generative adversarial networks can generate new samples with high quality that capture the intrinsic features of the original data, but not to simply memorize the training data. Furthermore, convolutional neural networks classification models show better classification performance than traditional machine learning models. And the performance of all these classification models is indeed improved to the different extent via the generative adversarial networks-based data augment. In addition, weather classification model plays a significant role in determining the most suitable and precise day-ahead photovoltaic power forecasting model with high efficiency.