Experimental verification of evolutionary estimation algorithms for radioactive source localisation

This paper considers localisation of point sources of gamma radiation using dose rate measurements. Binary and continuous genetic algorithms (GA) were used to implement maximum likelihood estimation (MLE) of position and strength of point radiation sources. MLE was achieved by minimising the objective function which computes the negative log likelihood. Real experimental data collected during a DSTO-conducted field trial were used to test and verify the performance of the algorithms. The performance of GA-based implementation was compared to an implementation that uses MATLAB built-in routine fminsearch. Source parameters estimated by the algorithms were also compared to the theoretical bounds obtained using Cramer-Rao bound (CRB) analysis which quantifies the accuracy with which it is possible to localise the source and estimate its strength. All three implementations localised a single point source well, nearly approaching the CRB. Reasonable position estimates were achieved for two and three source cases, but the source strength estimates were found to have much larger RMS errors than what is predicted by the CRB. While the GA-based implementations took longer to converge compared to the fminsearch algorithm, they encountered fewer divergent runs than the latter algorithm.