Development of Fast-Running Simulation Methodology Using Neural Networks for Load Follow Operation
Journal Article
·
· Nuclear Science and Engineering
OSTI ID:20804778
- Korea Atomic Energy Research Institute (Korea, Republic of)
- Seoul National University (Korea, Republic of)
- Yonsei University (Korea, Republic of)
A new fast-running analytic model has been developed for analyzing the load follow operation. The new model was based on the neural network theory, which has the capability of modeling the input/output relationships of a nonlinear system. The new model is made up of two error back-propagation neural networks and procedures to calculate core parameters, such as the distributions and density of xenon in a quasi-steady-state core like load follow operation. One neural network is designed to retrieve the axial offset of power distribution, and the other is for reactivity corresponding to a given core condition. The training data sets for learning the neural networks in the new model are generated with a three-dimensional nodal code and, also, the measured data of the first-day test of load follow operation. Using the new model, the simulation results of the 5-day load follow test in a pressurized water reactor show a good agreement between the simulation data and the actual measured data. Required computing time for simulating a load follow operation is comparable to that of a fast-running lumped model. Moreover, the new model does not require additional engineering factors to compensate for the difference between the actual measurements and analysis results because the neural network has the inherent learning capability of neural networks to new situations.
- OSTI ID:
- 20804778
- Journal Information:
- Nuclear Science and Engineering, Journal Name: Nuclear Science and Engineering Journal Issue: 1 Vol. 141; ISSN NSENAO; ISSN 0029-5639
- Country of Publication:
- United States
- Language:
- English
Similar Records
Neurocontrol of Pressurized Water Reactors in Load-Follow Operations
Application of an artificial neural network to reactor core analysis
Neural network: What it can do for petroleum engineers
Journal Article
·
Thu Dec 14 23:00:00 EST 2000
· Nuclear Technology
·
OSTI ID:20822227
Application of an artificial neural network to reactor core analysis
Journal Article
·
Sat Dec 30 23:00:00 EST 1995
· Transactions of the American Nuclear Society
·
OSTI ID:186687
Neural network: What it can do for petroleum engineers
Journal Article
·
Sat Dec 31 23:00:00 EST 1994
· JPT, Journal of Petroleum Technology; (United States)
·
OSTI ID:6683910