Loading...

Predication of Unmeasurable Parameters of the Reactor Core by Soft Computing Methods at the Transient State

Moradi, Milad | 2024

0 Viewed
  1. Type of Document: M.Sc. Thesis
  2. Language: Farsi
  3. Document No: 57492 (46)
  4. University: Sharif University of Technology
  5. Department: Energy Engineering
  6. Advisor(s): Ghaffari, Mohsen
  7. Abstract:
  8. The parameters within the reactor core can be categorized into two distinct groups of measurable parameters and unmeasurable parameters. The determination of unmeasurable parameters, such as void fraction and critical heat flux, plays a pivotal and fundamental role in predicting the occurrence of accidents and emergency situations within the reactor. The utilization of deep neural networks represents one of the methods for accurately and reliably estimating these parameters within future temporal intervals. Such estimations facilitate the implementation of necessary measures to prevent accidents or mitigate their consequences. In this study, we employ three deep neural network models namely LSTM, TFT, and NBEATS. Our objective is to estimate the void fraction within the reactor core. We explore three distinct approaches for neural network training: without covariates, using past covariates and using future covariates. Our findings reveal that the LSTM neural network, when trained with future covariates namely pressure, temperature, water velocity and steam velocity yields the lowest error. However, it is important to note that when these parameters occur, the void fraction also occurred. Consequently, using them as future covariates may not be suitable. In this research, we propose an approach that employ past covariates namely, pressure, temperature, steam velocity, and water velocity to training neural network model. Specifically, we utilize these past covariates to estimate the void fraction for future temporal intervals within the reactor core. Utilizing the TFT neural network, we achieved a favorable forecast for the upcoming 4 temporal intervals
  9. Keywords:
  10. Long Short Term Memory (LSTM) ; Unmeasurable Parameters ; Void Fraction ; Neural Basis Expansion Analysis For Interpretable Time Series (NBEATS) ; Temporal Fusion Transformer (TFT) ; Critical Heat Flux ; Deep Neural Networks

 Digital Object List

 Bookmark

No TOC