Abstract
The volatility of the wind generation reduces the profits of wind generators as a consequence of the differences between the real wind generation and the wind power offered in the electricity markets. This paper presents two models: i) wind and generic storage system offering without a physical connection and ii) wind and generic storage system offering with a physical connection to mitigate the wind positive imbalances (excess of the wind generation with respect to the wind power offered). The objective of the models is to maximize the expected profit of selling the energy in the day-ahead market, where the energy can come from the wind power and the storage system. Moreover, the wind power imbalance is penalized in the balancing market reducing the profits. The problems are modeled using stochastic mixed integer linear programming. A case study of a week (168 hours) is simulated to evaluate the models. After the simulations, the results are discussed and a summary of the main conclusions are presented.

Abstract
The new concept development is a critical stage of the innovation process that can be seen as a new knowledge creation process. This paper presents a new approach and a software tool for a collaborative new concept development. Our approach considers Collaborative Innovation Networks as ecosystems for new knowledge creation and integration, and Web Mashups as supporting platforms for the development of virtual co-learning and knowledge co-creation environments. The achieved results confirm the utility and efficacy of the software tool and allow foreseeing its suitability for use in educational contexts.

Abstract
The liberalization of energy markets has imposed several modifications in the electricity market environment. The paradigm of monopoly market ceased to exist, and new models have been put into practice. The new models have increased the incentive on competitiveness, making market players struggle to achieve the best outcomes out of market participation. Producers aim at reaching the maximum profit on the sale of energy, while consumers try to minimize their spending on electrical energy. The proposed methodology considers the optimization of players' participation in multiple market opportunities. Reference prices that are expected in each market type at each moment are achieved through the application of neural networks. Using the forecasted prices, the proposed portfolio optimization method allocates the sale and purchase of electrical energy to different markets throughout the time, with the aim at achieving the most advantageous participation profile. A particle swarm approach is used to reduce the execution time while guaranteeing the minimum degradation of the results. Results of the swarm methodology are compared to those of a deterministic approach, using real data from the Iberian electricity market - MIBEL.

Abstract

Abstract
In this paper, we investigate normal and nondegenerate forms of the maximum principle for optimal control problems with state constraints. We propose new constraint qualifications guaranteeing nondegeneracy and normality that have to be checked on smaller sets of points of an optimal trajectory than those in known sufficient conditions. In fact, the constraint qualifications proposed impose the existence of an inward pointing velocity just on the instants of time for which the optimal trajectory has an outward pointing velocity.

Abstract
Software reconfigurability became increasingly relevant to the architectural process due to the crescent dependency of modern societies on reliable and adaptable systems. Such systems are supposed to adapt themselves to surrounding environmental changes with minimal service disruption, if any. This paper introduces an engine that statically applies reconfigurations to (formal) models of software architectures. Reconfigurations are specified using a domain specific language-ReCooPLa-which targets the manipulation of software coordination structures, typically used in service-oriented architectures (soa). The engine is responsible for the compilation of ReCooPLa instances and their application to the relevant coordination structures. The resulting configurations are amenable to formal analysis of qualitative and quantitative (probabilistic) properties.