Optimization of Shielding- Collimator Parameters for ING-27 Neutron Generator Using MCNP5

Hegazy, Aya Hamdy; Skoy, V. R.; Hossny, K.

doi:10.1051/epjconf/201817702003

Cited by 4 publications

(4 citation statements)

References 4 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…This is because reinforcement learning methods, especially the off-policy ones, rely on previous experiences during training. These advantages do allow for more stability in deploying DRL models in critical applications, such as nuclear engineering [35][36][37][38].…”

Section: Discussionmentioning

confidence: 99%

Refined Continuous Control of DDPG Actors via Parametrised Activation

Hossny¹,

Iskander²,

Attia³

et al. 2021

View full text Add to dashboard Cite

Continuous action spaces impose a serious challenge for reinforcement learning agents. While several off-policy reinforcement learning algorithms provide a universal solution to continuous control problems, the real challenge lies in the fact that different actuators feature different response functions due to wear and tear (in mechanical systems) and fatigue (in biomechanical systems). In this paper, we propose enhancing the actor-critic reinforcement learning agents by parameterising the final layer in the actor network. This layer produces the actions to accommodate the behaviour discrepancy of different actuators under different load conditions during interaction with the environment. To achieve this, the actor is trained to learn the tuning parameter controlling the activation layer (e.g., Tanh and Sigmoid). The learned parameters are then used to create tailored activation functions for each actuator. We ran experiments on three OpenAI Gym environments, i.e., Pendulum-v0, LunarLanderContinuous-v2, and BipedalWalker-v2. Results showed an average of 23.15% and 33.80% increase in total episode reward of the LunarLanderContinuous-v2 and BipedalWalker-v2 environments, respectively. There was no apparent improvement in Pendulum-v0 environment but the proposed method produces a more stable actuation signal compared to the state-of-the-art method. The proposed method allows the reinforcement learning actor to produce more robust actions that accommodate the discrepancy in the actuators’ response functions. This is particularly useful for real life scenarios where actuators exhibit different response functions depending on the load and the interaction with the environment. This also simplifies the transfer learning problem by fine-tuning the parameterised activation layers instead of retraining the entire policy every time an actuator is replaced. Finally, the proposed method would allow better accommodation to biological actuators (e.g., muscles) in biomechanical systems.

show abstract

Section: Discussionmentioning

confidence: 99%

Refined Continuous Control of DDPG Actors via Parametrised Activation

Hossny¹,

Iskander²,

Attia³

et al. 2021

View full text Add to dashboard Cite

show abstract

“…The simulated setup is replicating the Romasha experimental setup located in Frank Laboratory at the Joint Institute for Nuclear Research (JINR) as demonstrated in Fig. 3 31 . The Romasha setup consisted of an ING-27 D-T neutron generator that generates 14.1 meV neutrons, six iron sheets collimator, and Ten BGO detectors located in a semicircle of 30 cm radius, as demonstrated in Fig.…”

Section: Data Generationmentioning

confidence: 99%

“…3. Dimensions of the setup are listed in Supplementary Table 2 22,31 . The gamma rays emitted due to neutron interactions with the sample travel in different directions.…”

Section: Data Generationmentioning

confidence: 99%

Detecting shielded explosives by coupling prompt gamma neutron activation analysis and deep neural networks

Hossny

Hossny²,

Magdi

et al. 2020

Sci Rep

Self Cite

View full text Add to dashboard Cite

Prompt Gamma Neutron Activation Analysis is a nuclear-based technique that can be used in explosives detection. It relies on bombarding unknown samples with neutrons emitted from a neutron source. These neutrons interact with the sample nuclei emitting the gamma spectrum with peaks at specific energies, which are considered a fingerprint for the sample composition. Analyzing these peaks heights will give information about the unknown sample material composition. Shielding the sample from gamma rays or neutrons will affect the gamma spectrum obtained to be analyzed, providing a false indication about the sample constituents, especially when the shield is unknown. Here we show how using deep neural networks can solve the shielding drawback associated with the prompt gamma neutron activation analysis technique in explosives detection. We found that the introduced end-to-end framework was capable of differentiating between explosive and non-explosive hydrocarbons with accuracy of 95% for the previously included explosives in the model development data set. It was also, capable of generalizing with accuracy 80% over the explosives which were not included in the model development data set. Our results show that coupling prompt gamma neutron activation analysis with deep neural networks has a good potential for high accuracy explosives detection regardless of the shield presence.

show abstract

“…Other researchers focused on the implementation of deep neural networks in different applications such as developing deep neural networks (DNNs) capable of helping in optimizing the radiotherapy treatment plan in the Oncology field of study [14]. Others worked on developing interconnected regressors and a classifier to differentiate between shielded explosive and non-explosive hydrocarbons using the signals obtained from the prompt gamma neutron activation analysis technique (PGNAA) based on the ROMASHA setup located in JINR, Russia [15,16]. Also, Hosny et al worked on developing a pipeline of interconnected K-nearest neighbor (KNN) regressors and a decision tree classifier capable of discriminating explosives from non-explosive hydrocarbons when the investigated sample was not shielded [17].…”

Section: Introductionmentioning

confidence: 99%

Determination of the full energy peak efficiency of HPGe detector for varying soil compositions and Marinelli beaker dimensions using Artificial Neural Networks

Hossny

Salama²,

Mageed³

et al. 2022

J. Inst.

Self Cite

View full text Add to dashboard Cite

Natural radioactivity measurement is studying the radiation emitted naturally from any nature's components such as air, soil, water, etc. This field of study plays an important role in public health and safety management. Scientists and engineers can calculate the radiation doses received by inhabitants of a certain area by measuring the natural radioactivity in the same area. One of the main issues facing the standard measurement procedure is determining the appropriate volume of the sample to be taken from the area under investigation. The volume of the investigated sample affects the detection efficiency of the HPGe detector due to the self-shielding effect. In this work we show how using artificial neural networks can help in choosing the proper dimensions of the Marinelli beaker to be used in the detection process. We used the various compositions of soil as a case study. Results showed that artificial neural networks were capable of finding the relation between soil composition, Marinelli beaker dimensions, and gamma energy as inputs and the HPGe detection efficiency as output. The developed model's regression was 99% compared to the detection efficiency calculated using MCNP. This will help to have a better prediction of the detection efficiency of natural radioactivity measurements.

show abstract

Optimization of Shielding- Collimator Parameters for ING-27 Neutron Generator Using MCNP5

Cited by 4 publications

References 4 publications

Refined Continuous Control of DDPG Actors via Parametrised Activation

Refined Continuous Control of DDPG Actors via Parametrised Activation

Detecting shielded explosives by coupling prompt gamma neutron activation analysis and deep neural networks

Determination of the full energy peak efficiency of HPGe detector for varying soil compositions and Marinelli beaker dimensions using Artificial Neural Networks

Contact Info

Product

Resources

About