Loading...

Development of a Validation and Calibration Algorithm for Thermohydraulic Sensors of Bushehr NPP First Circuit Using Neural Networks

Ebrahimzadeh, Alireza | 2021

175 Viewed
  1. Type of Document: M.Sc. Thesis
  2. Language: Farsi
  3. Document No: 54500 (46)
  4. University: Sharif University of Technology
  5. Department: Energy Engineering
  6. Advisor(s): Ghaffari, Mohsen; Moshkbar-Bakhshayesh, Khalil
  7. Abstract:
  8. Sensors are one of the most vital instruments in Nuclear Power Plants (NPP), and operators and safety systems monitor various parts of the NPP and control transients by analyzing the values reported by the sensors. Failure to detect malfunctions or anomalies in them would lead to catastrophic consequences. A new approach based on thermo-hydraulic simulation by RELAP5 code and Feed-Forward Neural Networks (FFNN) is given to detect faulty sensors and estimate their correct value which are two main objectives of the current study. This approach consists of two main parts; The first part, Fault Detection Hyper Block (FDHB), responsible for detecting faulty sensors, and the second part, Estimation Hyper Block (EHB), responsible for estimating the correct value of faulty sensors. The core of sensor detection and estimation in FDHB and EHB is FFNNs. Hence, to design an efficient neural net to achieve the objectives of this research, seven feature selectors (i.e., Information gain, ReliefF, F-regression, mRMR, Plus-L Minus-R, GA, and PSO), three sigmoid activation functions (i.e., Logistic, Tanh and Elliot), and three training algorithms (i.e., Levenberg–Marquardt (LM), Bayesian Regularization (BR) and Scaled Conjugate Gradient (SCG)) have been comprehensively compared and evaluated. The required data have been obtained by simulating LOFA and SBLOCA transients in the RELAP5 code for the Bushehr Nuclear Power Plant (BNPP). The main advantage of this approach over other proposed approaches is that with the failure of more than one sensor, the detection of other sensors is not completely disrupted, and are monitored continually and independently
  9. Keywords:
  10. Neural Network ; Feature Selection ; Estimation ; Identification ; Sensors ; Detection ; Nuclear Power Plants ; RELAP5 Code

 Digital Object List

 Bookmark

No TOC