Abstract
This paper presents a concept for building up a distributed monitoring and diagnosis system for complex industrial applications. For this purpose, a hierarchical organized model with distributed, cooperating agents was developed. The hierarchical aspect guarantees a predictable behaviour of the system with a high performance and the flexibility of the system is ensured by the federal distribution (Bongaerts 1998). By using this approach, a modular component diagnosis and monitoring (CDM) system is realized that enables the integration of legacy monitoring and diagnostic tools, specific to the application area. Universal applicable mechanisms were found to perform diagnostic processes and to improve the quality of a diagnosis by handling different diagnostic mechanisms in parallel and by applying conflict resolution algorithms. This software architecture for monitoring and diagnosis was developed by the University of Karlsruhe in cooperation with three industrial partners and one research institute within the framework of the EU Esprit Program: 'DIAMOND: DIstributed Architecture for MONitoring and Diagnosis' (DIAMOND 2002).

Abstract

Abstract
It is well known that optimization problems for the decision-making process in real environments should consider uncertainty to attain robust solutions. Although this uncertainty has been usually modelled using probability theory, assuming a random origin, possibility theory has emerged as an alternative uncertainty model when statistical information is not available, or when imprecision and vagueness have to be considered. This article proposes two different criteria to obtain robust solutions for linear optimization problems when the objective function coefficients are modelled with possibility distributions. To do so, chance constrained programming is used, leading to equivalent crisp optimization problems, which can be solved by commercial optimization software. A simple case example is presented to illustrate the use of the proposed methodology.

Abstract
Game theory has traditionally used real-valued utility functions in decision-making problems. However, the real information available to assess these utility functions is normally uncertain, suggesting the use of uncertainty distributions for a more realistic modelling. In this sense, utilities results or pay-offs have been normally modelled with probability distributions, assuming random uncertainty. However, when statistical information is unavailable, probability may not be the most adequate paradigm, and can lead to very large execution times when some real complex problems are addressed. In this article possibility distributions are used to model the uncertainty of utility functions when the strategies are probability distributions (mixed strategies) over a set of original and discrete strategies (pure strategies). Two dual approaches to solve the resulting non-cooperative fuzzy games are proposed: modelling players' risk aversion, and thus providing realistic conservative strategies. Two examples show the robustness of the strategies obtained with the proposed approaches.

Abstract
It is widely known and accepted that Nash equilibrium suitably models agents' behavior in electricity markets, since it is coherent with the common sense of their simultaneous profits maximisation. In the literature, these approaches are usually addressed using deterministic representations, despite the fact that electricity markets are highly conditioned by the uncertainty in demand or in agents' bidding strategies. Only some equilibrium-modelling approaches under uncertainty can be found in the literature, most of them using probability distributions. However, probability approaches may lead to very complex formulations and generally require restrictive assumptions (such as normality or independence) that can hardly be verified in real complex problems. A conjectured-price-response equilibrium model that uses LR-possibility distributions to represent the uncertainty of the residual demand curves faced by the participant agents is proposed. Modelling the risk-aversion attitudes of the agents, the resulting possibilistic equilibrium is transformed into a simplified deterministic one, which is solved with a new globally convergent algorithm for variational inequalities problems. Some interesting results for a real-size electricity system show the robustness of this new approach when compared with other risk-neutral approaches.

Abstract
MORSE is a set of tools developed for the Spanish electrical utility Endesa, for long term simulation and strategic analysis of the Spanish electricity sector. MORSE has three different types of modules: forecasting modules, to model the main businesses of the sector, data preparation modules, to define the sector structure, to import data from other Endesa planning tools and to prepare it for the forecasting modules, and analysis and reports modules, to prepare simulation scenarios to be feed to the forecasting modules, execute them, and collect and organize their outputs in appropriate reports. The pool simulation model of the system, EQUITEC, is a Conjectured Supply Function Equilibrium, where generation companies are represented at a technology level, and conjectures are endogenously computed under hypothesis of local linearity near the equilibrium point. The inclusion in EQUITEC of network and zonal constraints is currently being analyzed. © 2010 IEEE.