International Conference on Mathematics and Computational Methods Applied to Nuclear Science and Engineering (M&C 2011) Rio de Janeiro, RJ, Brazil, May 8-12, 2011, on CD-ROM, Latin American Section (LAS) / American Nuclear Society (ANS) ISBN 978-85-63688-00-2 ACCIDENT IDENTIFICATION SYSTEM WITH AUTOMATIC DETECTION OF ABNORMAL CONDITION USING QUANTUM COMPUTATION Andressa dos Santos Nicolau ∗ , and Roberto Schirru Federal University of Rio de Janeiro, andressa@lmp.ufrj.br ; schirru@lmp.ufrj.br Alan Miranda Monteiro de Lima Nuclear Engineering Program – COPPE/UFRJ Ilha do Fundão s/n, Zip Code 21.945-970, Rio de Janeiro, Brazil ABSTRACT Transient identification systems have been proposed in order to maintain the plant operating in safe conditions and help operators in make decisions in emergency short time interval with maximum certainty associated. This article presents a system, time independent and without the use of an event that can be used as a starting point for t = 0 (reactor scram, for instance), for transient/accident identification of a pressurized water nuclear reactor (PWR). The model was developed in order to be able to recognize the normal condition and three accidents of the design basis list of the Nulear Power Plant Angra 2, postulated in the Final Safety Analysis Report (FSAR). Were used several sets of process variables in order to establish a minimum set of variables considered necessary and sufficient. The optimization step of the identification algorithm is based upon the paradigm of Quantum Computing. In this case, the optimization metaheuristic Quantum Inspired Evolutionary Algorithm (QEA) was implemented and works as a data mining tool. The results obtained with the QEA without the time variable are compatible to the techniques in the reference literature, for the transient identification problem, with less computational effort (number of evaluations). This system allows a solution that approximates the ideal solution, the Voronoi Vectors with only one partition for the classes of accidents with robustness. Key Words: Diagnostic systems, Nuclear power plant, Quantum computing, Artificial Inteligence. 1. INTRODUCTION The safety philosophy of nuclear reactors is directed toward that nuclear power plants are designed, built and operated to the highest quality standards and conditions that have high reliability. The prevailing thinking is that in an environment of zero tolerance, you can always improve safety.