Using genetic algorithms to calibrate the user-defined parameters of IIST model for SBLOCA analysis

Abstract The thermal–hydraulic system codes, i.e., TRACE, have been designed to predict, investigate, and simulate nuclear reactor transients and accidents. Implementing relevant correlations, these codes are able to represent important phenomena such as two-phase flow, critical flow, and countercurrent flow. Furthermore, the thermal–hydraulic system codes permit users to modify the coefficients corresponding to the correlations, providing a certain degree of freedom to calibrate the numerical results, i.e., peak cladding temperature. These coefficients are known as user-defined parameters (UDPs). Practically, defining a series of UDPs is complex, highly relied on expert opinions and engineering experiences. This study proposes another approach – the genetic algorithms (GAs), providing rigorous procedures and mitigating human judgments and mistakes, to calibrate the UDPs of important correlations for a 2% small break loss of coolant accident (SBLOCA). The TRACE IIST model was employed as a case study to demonstrate the capability of GAs. The UDPs were evolved by GAs to reduce the deviations between TRACE results and IIST experimental data.

[1]  C. H. Lee,et al.  Investigation of mid-loop operation with loss of RHR at INER integral system test (IIST) facility , 1996 .

[2]  Tay-Jian Liu,et al.  Using an IIST SBLOCA Experiment to Assess RELAP5/MOD3.2 , 1999 .

[3]  Enrico Zio,et al.  Genetic Algorithms for Estimating Effective Parameters in a Lumped Reactor Model for Reactivity Predictions , 2001 .

[4]  Enrico Zio,et al.  Using genetic algorithms for calibrating simplified models of nuclear reactor dynamics , 2004 .

[5]  Kalyanmoy Deb,et al.  Multi-objective optimization using evolutionary algorithms , 2001, Wiley-Interscience series in systems and optimization.

[6]  S. C. Chetal,et al.  Steam condenser optimization using Real-parameter Genetic Algorithm for Prototype Fast Breeder Reactor , 2011 .

[7]  Lance L. C. Wang,et al.  ICONE11-36446 ASSESSMENT OF THE RELAP5-3D CODE Against An IIST 0.5% COLD LEG SBLOCA EXPERIMENT WITH PASSIVE SAFETY INJECTION , 2003 .

[8]  Xue Dazhi,et al.  A genetic algorithm solution for a nuclear power plant risk–cost maintenance model , 2004 .

[9]  Igor Pioro,et al.  Appropriate thermodynamic cycles to be used in future pressure-channel supercritical water-cooled nuclear power plants , 2012 .

[10]  Jong-Rong Wang,et al.  Assessment of TRACE CCFL Model with SBLOCA Experiment of IIST Facility , 2012 .

[11]  Jasbir S. Arora,et al.  Survey of multi-objective optimization methods for engineering , 2004 .

[12]  Sofía Carlos,et al.  Parameter estimation in thermalhydraulic models using the multidirectional search method , 2003 .

[13]  Luca Podofillini,et al.  Model parameters estimation and sensitivity by genetic algorithms , 2003 .