Fuzzy logic for signal prediction in nuclear systems

Prediction of Automatic Scram during Abnormal Conditions of Nuclear Power Plants Based on Long Short-Term Memory (LSTM) and Dropout

A deep-learning model was proposed for predicting the remaining time to automatic scram during abnormal conditions of nuclear power plants (NPPs) based on long short-term memory (LSTM) and dropout. The proposed model was trained by simulated condition data of abnormal conditions; the input of the model was the deviation of the monitoring parameters from the normal operating state, and the output was the remaining time from the current moment to the upcoming reactor trip. The predicted remaining time to the reactor trip decreases with the development of abnormal conditions; thus, the output of the proposed model generates a predicted countdown to the reactor trip. The proposed prediction model showed better prediction performance than the Elman neural network model in the experiments but encountered an overfitting problem for testing data containing noise. Therefore, dropout was applied to further improve the generalization ability of the prediction model based on LSTM. The proposed automatic scram prediction model can provide NPP operators with an alert to the automatic scram during abnormal conditions.

On-line process monitoring during transient operations using weighted distance Auto Associative Bilateral Kernel Regression

In this paper, a new data-driven auto associative bilateral kernel regression (AABKR) method based on weighted distance is proposed for the on-line monitoring of transient process operations. A bilateral approach to the kernel regression formulates a representative model that uses both the spatial and temporal information in the data, and a new weighted-distance algorithm captures temporal information. Moreover, an adaptive approach is proposed to dynamically compensate for faulty process inputs in the bilateral kernel evaluations, providing a robust model with little spillover. The proposed weighted-distance AABKR is first implemented using numerical process examples and then applied to the transient start-up operation of a nuclear power plant. Monte Carlo simulation results are provided by randomly assigning fault sensors and fault magnitudes. The results demonstrate the feasibility and efficiency of the proposed method.