Detection of radiological dispersal devices (RDDs) is important because of their potential for destruction and psychological impact on the affected population. These devices leave a clear trace which can be followed when using appropriate detection devices. Geiger counter devices provide data regarding the radiation intensity. However, this is not enough information to pinpoint a radiation source. Neither can this information be directly used to classify the radiation source. We describe a method using symbiotic simulation which can be used to classify and localise a radiation source given accurate measurements of radiation intensities at reference points and a detailed model of the environment. Initial classification and localisation, as well as continuous tracking of a moving radiation source, is considered. The effects of a measurement error and a model error are investigated.
展开▼