Building a bayesian framework to analyse measurements from the ligne d´intefration laser facility

The LIL facility is the small-scale prototype of the Laser MegaJoule (LMJ) that is one of the two high power laser facilities for nuclear fusion by inertial confinement. Several diagnostics have been set up to characterise both laser performances and laser-matter interactions. Therefore, we wish to describe the state of a physical system knowing noisy measurements, by using a probabilistic approach. Following Clapp et al. (2000), we have set up a simulated annealed particle filter that determine the posterior probability density function given measurement, calibration data and a priori knowledge on the analysed system. This implementation has been applied to the characterization of focal spot intensity