Nicolas Paul

On-line restoration for turbulence degraded video in nuclear power plant reactors

This article deals with video inspection of nuclear plant reactor after fuel reloading. During these underwater inspections, the fuel assemblies’ heat generates turbulence effect that sensitively degrades the video quality. An on-line restoration algorithm is proposed here. It consists of two steps. A temporal infinite impulse response filter is first used to get a stabilized but blurry video. A second spatial Wiener deconvolution filter is then used to estimate the video which would have been observed without turbulence. This second filter is based on a probabilistic model of the turbulence impact on the observed video. An on-line prototype, based on this algorithm and its straightforward extension to a moving camera (translation), has been successfully tested on several power plants.

Low complexity algorithm for the decoding of convolutional codes of any rate

It is well known that convolutional codes can be optimally decoded by using the Viterbi algorithm (VA). A decoding technique where the VA is applied to identify the error vector rather than the information message is proposed. We previously focused on convolutional coders of rate 1/2 . The method to codes of any rate is generalized and shows that, with the proposed type of decoding, the exhaustive computation of a vast majority of state to state iterations is unnecessary. Hence, performance close to optimum is achievable with a significant reduction of complexity. The higher the SNR, the greater the improvement for reduction in complexity. For instance, for SNR greater than 3 dB, a five fold reduction in complexity for the computation of ACS (add compare select) is achieved.

Estimation of a generic model for a fleet of machines

In this paper, we propose a method to identify a generic model for a machine which is a member of a fleet of identical machines. These latest may work in the same or in different operating conditions. Firstly, the method consists in estimating a linear model for each machine independently from other machines, based on the data collected on the machine itself. Secondly, the common part of the models is identified. For that purpose, a resemblance criterion is used to determine identical coefficients over all the models. New models are then generated under some equality constraints on these parameters. The last step of the method consists in validating the choice made about the common part and hence validating the new models. This is realized by a statistical test based on a comparison between new and old estimates. An academic example is finally presented to illustrate the results of the method.