Complex Lyapunov Exponents from Short and Noisy Sets of Data. Application to Stability Analysis of BWRs

The dynamics reconstruction techniques have been applied to systems as BWRs with a big amount of noise. The success of this methodology was limited due to the noise in the signals. Recently. new techniques have been introduced for short and noisy data sets based on a global fit of the signal by means of orthonormal polynomials. In this paper, we revisit these ideas in order to adapt them for the analysis of the neutronic power signals to characterize the stability regime of BWR reactors. To check the performance of the methodology. we have analyzed simulated noisy signals. observing that the method works well. even with a big amount of noise. Also. we have analyzed experimental signals from Ringhals I BWR. In this case. the reconstructed phase space for the system is not very good. A modal decomposition treatment for the signals is proposed producing signals with better behaviour.