Abstract
Insular power systems are characterized by their isolated geographical location, which makes their interconnection with other power systems a challenging task. Moreover, these islands have important renewable resources that allow the reduction of generation costs and greenhouse gas emissions (GHE). To guaranty the quality, flexibility and robustness of the electrical framework, the representation of renewable power forecasting error by scenario generation or even the implementation of demand response tools have been adopted. In this paper, the failure events of a specific unit are considered according to its capacity. Then, using the forced outage rate, the probability of each failure event is computed. Results of energy not supplied and fuel consumption cost are determined by applying probabilistic concepts, while the final results are obtained by fitting and evaluating a nonlinear trend line carried out using the previous results, resulting in a proficient computational tool compared with classical ones.

Abstract
Although wind power generation has extended a maturity in technology, there are still many concerns regarding the optimal support of regulatory bodies for renewable resources. In this context, the regulatory body should form a market structure or consider market rules and regulations to not only attract investors to renewable power plants, but also provide an efficient market that reflects a safe and clear competition environment. In this paper, an agent-based game-theoretic model is developed to investigate the electricity market behavior under oligopoly circumstances. The proposed model reveals the potential of collusive and strategic behavior of market participants. By employing the proposed model, impacts of different supportive schemes on the behavior of the wind power producer and conventional thermal units are investigated. According to the results obtained, if the regulatory bodies do not consider strategic collusion of market participants, adverse consequences for wind power producers might happen in the long-term horizon.

Abstract
Participation of consumers in Demand Response (DR) programs improves system stability and reliability as well as market efficiency. Retailers and distributors purchase DR to advance business and system reliability, respectively. Meanwhile, large consumers, Distribution System Operators (DSOs), Load Service Entities (LSEs), and DR aggregators sell DR to increase their own profits. In this context, DR aggregators are key elements of power systems that enhance the participation of consumers in electricity markets. These market participants can negotiate their aggregated DR with other market players in Demand Response eXchange (DRX) markets, and participate in the energy and ancillary service markets. Hence, this paper proposes a stochastic model to optimize the performance of a DR aggregator to take part in the day-ahead energy, ancillary services and intraday DRX markets. In order to mitigate the negative impacts of uncertainties, Conditional Value at Risk (CVaR) is also incorporated to the proposed model. Numerical studies indicate that the proposed model for DR aggregator can arrange its offering/bidding strategies to participate in the mentioned markets simultaneously.

Abstract
Development of distributed energy resources introduces high level of interdependency and the need for integrated models in a multienergy system (MES). Moreover, highlighting environmental aspects facilitates electrification in the transportation sector and integration of plug-in electric vehicles (PEVs). In this paper, aggregation of PEVs' batteries in parking lots (PL) is considered as a bulk electric storage in MES. The energy hub approach is employed for modeling MES considering PL. Due to the profitable behavior of PL in the reserve market, the energy hub model is modified to consider the reserve sources as ancillary services in the output energy vector. Moreover, the uncertain traffic pattern of PEVs' owners in PL is modeled by a stochastic approach. The numerical results demonstrate the proficiency of the proposed model, determining the changes in the behavior of other MESs elements in the presence of PL.

Abstract
This paper presents a new stochastic multi-layer agent-based decision making model to study the behavior of market participants in the future smart grid. In the agent-based model proposed, wholesale market players are modeled in the first layer. The players include renewable/sustainable power producers, optimizing the bidding/offering strategies to participate in the electricity markets. In the second layer, responsive customers include electric vehicle owners and consumers who participate in demand response programs, being modeled as independent agents. The objective of the responsive customers is to increase their benefit while retaining welfare. The interaction between market players in day-ahead and real-time markets is modeled using an incomplete information game theory algorithm. According to the high uncertainty of resources and customers' behavior, the model is developed using a stochastic framework. A case study containing wind power producers, aggregators and retailers providing demand response is considered to confirm the usefulness of the proposed multi-layer model.

Abstract
In this paper, a multi-period integrated framework is developed for generation expansion planning (GEP), transmission expansion planning (TEP), and natural gas grid expansion planning (NGGEP) problems for large-scale systems. New nodal generation requirements, new transmission lines, and natural gas (NG) pipelines are simultaneously obtained in a multi-period planning horizon. In addition, a new approach is proposed to compute NG load flow by considering grid compressors. In order to solve the large-scale mixed integer nonlinear problem, a framework is developed based on genetic algorithms. The proposed framework performance is investigated by applying it to a typical electric-NG combined grid. Moreover, in order to evaluate the effectiveness of the proposed framework for real-world systems, it has been applied to the Iranian power and NG system, including 98 power plants, 521 buses, 1060 transmission lines, and 92 NG pipelines. The results indicate that the proposed framework is applicable for large-scale and real-world systems.