Artificial Neural Networks (ANNs) are commonly used in place of expensive models to reduce the computational burden required for uncertainty quantification, reliability and sensitivity analyses. ANN with selected architecture is trained with the back-propagation algorithm from few data representatives of the input/output relationship of the underlying model of interest. However, different performing ANNs might be obtained with the same training data as a result of the random initialization of the weight parameters in each of the network, leading to an uncertainty in selecting the best performing ANN. On the other hand, using cross-validation to select the best performing ANN based on the ANN with the highest R value can lead to biassing in the prediction. This is as a result of the fact that the use of R cannot determine if the prediction made by ANN is biased. Additionally, R does not indicate if a model is adequate, as it is possible to have a low R for a good model and a high R for a bad model. Hence, in this paper, we propose an approach to improve the robustness of a prediction made by ANN. The approach is based on a systematic combination of identical trained ANNs, by coupling the Bayesian framework and model averaging. Additionally, the uncertainties of the robust prediction derived from the approach are quantified in terms of confidence intervals. To demonstrate the applicability of the proposed approach, two synthetic numerical examples are presented. Finally, the proposed approach is used to perform a reliability and sensitivity analyses on a process simulation model of a UK nuclear effluent treatment plant developed by National Nuclear Laboratory (NNL) and treated in this study as a black-box employing a set of training data as a test case. This model has been extensively validated against plant and experimental data and used to support the UK effluent discharge strategy.
Abstract. The dose distributions from proton pencil beam scanning were calculated by FLUKA, GEANT4, MCNP, and PHITS, in order to investigate their applicability in proton radiotherapy. The first studied case was the integrated depth dose curves (IDDCs), respectively from a 100 and a 226-MeV proton pencil beam impinging a water phantom. The calculated IDDCs agree with each other as long as each code employs 75 eV for the ionization potential of water. The second case considered a similar condition of the first case but with proton energies in a Gaussian distribution. The comparison to the measurement indicates the inter-code differences might not only due to different stopping power but also the nuclear physics models. How the physics parameter setting affect the computation time was also discussed. In the third case, the applicability of each code for pencil beam scanning was confirmed by delivering a uniform volumetric dose distribution based on the treatment plan, and the results showed general agreement between each codes, the treatment plan, and the measurement, except that some deviations were found in the penumbra region. This study has demonstrated that the selected codes are all capable of performing dose calculations for therapeutic scanning proton beams with proper physics settings.
High-energy neutrons (>10 MeV) contribute substantially to the dose fraction but result in only a small or negligible response in most conventional moderated-type neutron detectors. Neutron dosemeters used for radiation protection purpose are commonly calibrated with (252)Cf neutron sources and are used in various workplace. A workplace-specific correction factor is suggested. In this study, the effect of the neutron spectrum on the accuracy of dose measurements was investigated. A set of neutron spectra representing various neutron environments was selected to study the dose responses of a series of Bonner spheres, including standard and extended-range spheres. By comparing (252)Cf-calibrated dose responses with reference values based on fluence-to-dose conversion coefficients, this paper presents recommendations for neutron field characterisation and appropriate correction factors for responses of conventional neutron dosemeters used in environments with high-energy neutrons. The correction depends on the estimated percentage of high-energy neutrons in the spectrum or the ratio between the measured responses of two Bonner spheres (the 4P6_8 extended-range sphere versus the 6″ standard sphere).
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.