Year
1997
Abstract
Along with the new term \"data fusion\" have come several new or revived data analysis methods. Applications of data fusion include combining information from radiation sensors to assay an item containing nuclear material (a micro application) and combining quite different sensor data such as image and radiation, and expert opinion to judge whether a facility is operating as declared (a macro application). This paper focuses on data fusion for nondestructive assay (NDA) of special nuclear material. Some data fusion methods have promise in empirical or semi-empirical approaches to calibrating NDA methods. We review where some of these methods have been or could be used in NDA applications and indicate the expected improvement in assay uncertainty for a few cases. Some cases involve rather disparate data, for example, from neutron and gamma counters, but we focus on analysis of multiple sensors of the same type (neutron counters). We present analysis results for a subset of a large data set of 252Cf shuffler assays (via neutron counting) of large containers with known amounts of plutonium in known locations in known matrices.