Advanced Multi-Modeling of PWR Dynamics and Deep Learning based Computational Tool in SIMULINK and LabVIEW
Advanced DeepComputational Multi-Model PWR Neutronics
DOI:
https://doi.org/10.53560/PPASA(59-1)748Keywords:
Multi-Model PWR Dynamics, Deterministic Reactor Neutronics, Deep Stochastic Optimization, Artificial Intelligence, Hybrid SimulationAbstract
The reactivity monitoring, prediction, and investigation is the most important parameter to ensure the safety and reliable operation of a nuclear power plant. This parameter is gained further importance in Pressurized Water Reactor (PWR) due to more sophisticated reactivity insertion mechanisms and innovative reactor core fuel loading scheme. Based on deterministic internal and external dynamics and neutronics analysis of Advanced PWR, all the reactivity feedback effects such as Doppler effect, moderator effect, control rod effect, liquid boron effect and reactor poisons effect are investigated, modeled and stochastically optimized using deep artificial intelligence. Advance Pressurized Water Reactor (APWR) of 600 MWe rating (AP-600) is used as a reference reactor model and based on the dynamics of AP-600, an Advanced Pressurized Water Reactor Dynamics and Intelligent Stochastic Optimization based Deterministic Neutronics Simulation (APD-ISO-DNS) Code is developed in the hybrid SIMULINK and
LabVIEW environments. AP-600 reactor model is fine-tuned and adjusted for 300 MWe PWR (P-300) and 1070 MWe Advanced Chinese PWR (ACP-1000) using neutronics parameters and operational dynamic data of operating PWR nuclear power plants in Pakistan. Various load reduction transient experiments are conducted and dynamic transient simulations of P-300, AP-600 and ACP-1000 are evaluated in SIMULINK and in LabVIEW environments and found as per design basis.
References
M. Subekti, S. Bakhri, and G. R. Sunaryo.The simulator development for RDE reactor. Journal of Physics 962: (2018).
. M. Johnson, S. Lucas, and P. Tsvetkov. Modeling of reactor kinetics and dynamics.Idaho National Laboratory, Report Idaho Falls, Idaho 83415, U.S. Department of Energy, Canada (2010).
W. K. Lam. Advanced pressurized water reactor simulator. IAEA Workshop on NPP Simulators for Education, Bucharest, Romania (2006).
L. C. C. Po. PCTRAN/PWR. Report Montville, New Jersey 07045, IAEA Workshop on NPP Simulators for Education, Bucharest, Romania (2006).
L. C. C. Po. PC-based simulator for education in advanced nuclear power plant construction. International Symposium on the Peaceful Application of Nuclear Technology in the GCC Countries (2008).
A. S. Mollah. Education tool for simulation of safety and transient analysis of a pressurized water reactor. International Journal of Integrated Sciences & Technology 03: 01-10 (2018).
S. M. A. Ibrahim. Thermal hydraulic simulations of a PWR nuclear power plant. International Journal of Safety and Security 013 (01): 31-520 (2019).
A. H. Malik,, A. A. Memon, and M. R. Khan. Identification of nonlinear dynamics of nuclear power reactor using adaptive feedforward neural network. Proceedings of Pakistan Academy Sciences 47 (2): 111-120 (2010).
Z. Wenjie, Q. Jiang, Jinsen X. and T. Yu. A functional variable universe fuzzy PID controller for load following operation of PWR with the multiple model. Annals of Nuclear Energy140: 1-6 (2020).
S. U. E. Hakim, A. Abimanyu, and Sutanto. Simulator design of Kartini reactor based on LabVIEW. Journal forum Nukir 12 (1): 29-41 (2018).
L. A. Macedo, W. M. Torres, G. Sabundjian, D. A. Andrade, A. B. Junior, P. E. Umbehaun, T. N. Conti, R. N. Mesquita, P. H. F. Masotti, and G. Angelo. Development of a LabVIEW web-based simulator for RELAP. International Nuclear Atlantic Conference, Brazil: 01-13 (2011).
J. Park, J. Jang, H. Kim, .Choe, D. Yun, and P. Zhang.RAST-K V2-Three-Dimentional nodal diffusion code for pressurized water reactor core analysis. Energies 13: 01-21 (2020).
X. Cui, W. Zhang, Z. Tuske, and M. Picheny. Evolutionary stochastic gradient decent for optimization of deep neural networks. 32 Conference on Neural Information Processing Systems, Canada: (2018).
W. An, H. Wang, Q. Sun, J. Xu, Q. Dai, and L. Zhang.A PID controller approach for stochastic optimization of deep neural networks. IEEE / CVF Conference on Computer Vision Pattern Recognition: (2018).
H. Wang, Y. Luo, W. An, Q. Sun, J. Xu, and L. Zhang.PID controller-based stochastic optimization acceleration for deep neural networks. IEEE Transactions on Neural Networks and Learning Systems 31 (12): 1-10 (2020).
J. H. Horng. Hybrid MATLAB and LabVIEW with neural network to implement a SCADA system of AC servo motor. Advances in Engineering Software 39: 149-155 (2008).
