Study on Noise Detection of Sodium-water Reaction in the Steam Generator of the Sodium Cooled Fast Reactor Based on Long-and Short-term Memory
The sodium water reaction caused by the leakage of the heat transfer pipeline in the steam generator of the sodium cooled fast reactor is difficult to identify in the early stage and develops rapidly.In response to the above phenomenon,a steam generator small leakage signal recognition and detection method based on Long Short Term Memory(LSTM)network is proposed.The leakage signal observation data under the maximum background noise condition of the test bench is used as input to establish a sodium water reaction LSTM classification model,which realizes the recognition of small leakage signals under strong background noise operating conditions.The experimental results show that the proposed LSTM classification model has extremely high judgment accuracy.Compared with other classification methods,the proposed method has better accuracy and reliability in leakage judgment,verifying the above LSTM based classification model.The feasibility and effectiveness of the sodium-water reaction leakage detection method for steam generators using algorithms for small leakage detection under strong background noise conditions.
sodium cooled fast reactorsteam generatorsodium-water reactionlong-and short-term memorysignal classification
NETL
NSTL
万方数据
