A load-following operation in APR+ nuclear plants is necessary to reduce the need to adjust the boric acid concentration and to efficiently control the control rods for flexible operation. In particular, a disproportion in the axial flux distribution, which is normally caused by a load-following operation in a reactor core, causes xenon oscillation because the absorption cross-section of xenon is extremely large and its effects in a reactor are delayed by the iodine precursor. A model predictive control (MPC) method was used to design an automatic load-following controller for the integrated thermal power level and axial shape index (ASI) control for APR+ nuclear plants. Some tracking controllers employ the current tracking command only. On the other hand, the MPC can achieve better tracking performance because it considers future commands in addition to the current tracking command. The basic concept of the MPC is to solve an optimization problem for generating finite future control inputs at the current time and to implement as the current control input only the first control input among the solutions of the finite time steps. At the next time step, the procedure to solve the optimization problem is then repeated. The support vector regression (SVR) model that is used widely for function approximation problems is used to predict the future outputs based on previous inputs and outputs. In addition, a genetic algorithm is employed to minimize the objective function of a MPC control algorithm with multiple constraints. The power level and ASI are controlled by regulating the control banks and part-strength control banks together with an automatic adjustment of the boric acid concentration. The 3-dimensional MASTER code, which models APR+ nuclear plants, is interfaced to the proposed controller to confirm the performance of the controlling reactor power level and ASI. Numerical simulations showed that the proposed controller exhibits very fast tracking responses.

The departure from nucleate boiling ratio (DNBR) is one of the most critical parameters in the safety issues of a nuclear reactor. Most reactor core protection systems of current nuclear power plants calculate the minimum DNBR at a pseudo hot fuel rod position to prevent the departure from nucleate boiling (DNB). On the other hand, it gives rise to a more conservative result, which reduces the operating margin of nuclear power plants. In this paper, the axial DNBR distribution at the actual hot fuel rod position was predicted based on the support vector regression (SVR) model, which is a data-based method using a number of measured signals from the reactor coolant system. SVR models were developed using a learning data set and validated by an independent test data set. These models were applied to the first fuel cycle of the Yonggwang unit 3 nuclear power plant. The root mean square (RMS) error averaged for 13 axial locations of the hot rod was 0.87%. The SVR models estimate DNBR values more accurately at central parts that have relatively lower DNBR values, which are more important in terms of safety. This algorithm can predict the DNBR accurately at each time step and provide reliable protection and monitoring information for nuclear power plant (NPP) operation.

The diametral creep of pressure tubes (PTs) in CANDU (CANada Deuterium Uranium) reactors is one of the principal aging mechanisms governing the heat transfer and hydraulic degradation of the heat transport system (HTS). PT diametral creep leads to diametral expansion, which affects the thermal hydraulic characteristics of the coolant channels and the critical heat flux (CHF). The CHF is a major parameter determining the critical channel power (CCP), which is used in the trip setpoint calculations of regional overpower protection (ROP) systems. Therefore, it is essential to predict PT diametral creep in CANDU reactors. PT diametral creep is caused mainly by fast neutron irradiation, temperature and applied stress. The objective of this study was to develop a bundle position-wise linear model (BPLM) to predict PT diametral creep employing previously measured PT diameters and HTS operating conditions. The linear model was optimized using a genetic algorithm and was devised based on a bundle position because it is expected that each bundle position in a PT channel has inherent characteristics. The proposed BPLM for predicting PT diametral creep was confirmed using the operating data of the Wolsung nuclear power plant in Korea. The linear model was able to predict PT diametral creep accurately.

It is very important for operators to be informed of the departure from nucleate boiling ratio (DNBR) to prevent the fuel cladding from melting and causing a boiling crisis. Artificial intelligence methods such as neural networks and support vector regression (SVR) have extensively and successfully been applied to nonlinear function approximation. In this paper, fuzzy support vector regression (FSVR) combined with a fuzzy concept and SVR is presented to precisely predict the minimum DNBR by using the measured signals of a reactor coolant system, such as reactor power, reactor pressure, and control rod positions. Also, the prediction uncertainty for the predicted minimum DNBR is assessed. It is demonstrated that FSVR is accurate enough to be used in protection and monitoring algorithms for departure from nucleate boiling (DNB). Therefore, FSVR can be used to effectively monitor and predict the minimum DNBR in the reactor core.