Development of a three-dimensional model for sensor positioning

Abstract

In the field of radiation detection, detector placement is dependant on the type of radiation detector, end user application, required data fusion and complexity of the system. Homeland Security needs may require a variety of deployment specific configurations; examples include the optimum placement of detectors within a portal rack, or, the positioning of a distributed network of sensors at a particular traffic intersection. This paper details a computational model developed to enable the area of coverage provided by a number of detectors to be rapidly calculated. Starting from a two-dimensional analytical method, a random sequential deposition numerical three-dimensional model is described. By sitting the entire environment on a mesh, the total volume of detector coverage is calculated via a numerical dynamic approach. Comparing the area of detector coverage with the volume traversed by radiological material moving through the environment, estimates of the likelihood of detection as a function of detector coverage were determined. To improve on random deposition, space optimised sequential deposition methods are detailed. Comments are also included on the ability to couple deposition models with machine learning approaches, so enabling intelligent learning for optimised detector placement across a range of deployments