[Background]: The thermal stratification in the upper plenum of lead-bismuth fast reactor after emergency shutdown has a significant impact on the structural integrity of the reactor and the residual heat removal capacity of the natural circulation.The research on thermal stratification based on CFD method has the problems of large computational overhead and time-consuming

while the existing standard dynamic mode decomposition ( DMD ) method has poor forecasting results on thermal stratification. [Purpose]: In order to solve this problem

this paper proposes a thermal stratification model reduction method for the upper plenum of lead-bismuth fast reactor. [Methods]: Firstly

the high-precision full-order snapshot is obtained based on the CFD program FLUENT. Then

based on the truncated DMD

the time step samples are compressed according to the characteristic frequency

and the thermal stratification reduction model is constructed by combining the LSTM neural network. [Results] By comparing and analyzing the thermal stratification phenomenon with the CFD full-order thermal stratification simulation

the thermal stratification reduction model can better simulate the temperature distribution in the upper plenum and realize the rapid prediction of the thermal stratification phenomenon. [Conclusion] The efficiency of the thermal stratification model reduction method based on improved DMD and LSTM in the upper plenum of the lead-bismuth fast reactor proposed in this paper is significantly higher than that of the traditional thermal stratification calculation

which verifies that the thermal stratification model reduction method in this paper has certain application value.