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

<div>Prediction results for testing data containing artificial noise (max ±0.07) by sufficiently trained LSTM model with a different dropout rate: (a) no dropout; (b) dropout rate = 0.2; (c) dropout rate = 0.5.</div>

Science and Technology of Nuclear Installations

fig13

Figure 13

Figure 13: Prediction of Automatic Scram during Abnormal Conditions of Nuclear Power Plants Based on Long Short-Term Memory (LSTM) and Dropout 