Self Powered neutron detector based reactor flux estimation using multisensor particle filter

Abstract

In this paper, we propose an improved method of In-core neutron flux estimation using Inconel Self Powered Neutron Detectors (SPNDs). The method proposed here is based upon multi-sensor based particle filter utilizing a number of uniformly distributed SPNDs along with the out-core ionization chambers which measure the overall flux. The proposed method estimates the neutron flux of large non-linear core volume involving stiff non-linearities with non-Gaussian uncertainties both in the process and sensor model. We propose an improved particle filtering based data fusion algorithm on multi-sensor network for flux estimation under nonlinear non-Gaussian environment. The nonlinear system in our study is a large core nuclear reactor measured through in-core Self Powered Neutron Detectors. Many critical applications such as reactor protection and control rely upon the neutron flux information and thus make the reliability of data an utmost important. The point kinetic model based on neutron transport theory conveniently explains the dynamics of nuclear reactor. The state equation in point kinetic model is stiff nonlinear set of equations. The neutron flux in the large core, loosely coupled reactor are sensed by multiple sensors measuring point fluxes located at various locations inside the reactor core. The flux values are coupled to each other through diffusion equation. The coupling facilitates redundancy in the information. The multiple independent information about the localized flux peaking can be fused together to enhance the estimation accuracy to a great extent. In our work, we establish, observation model for the neutron flux sensor used for large core flux measurement i.e. Self Powered Neutron Detectoric document.