The rapid and accurate prediction for the temperature rise of gas insulated switchgear (GIS) is a necessary prerequisite for realizing its thermal health management. However

the complex boundary conditions on site make it difficult to keep the prediction consistent with the actual temperature rise. In order to obtain the transient temperature rise and heat transfer characteristics of three-phase enclosure GIS bus under the constraints of complex unknown heat dissipation boundary on site. In this paper

a three-phase enclosure GIS simulation model based on multiple-relaxation- time lattice Boltzmann method (MRT-LBM) and ensemble Kalman filter (EnKF) data assimilation method is proposed. Firstly

the multi-relaxation scheme is used to improve the numerical stability of the GIS model under high Rayleigh number (Ra). Then the EnKF algorithm is used to assimilate the MRT-LBM calculation results and the experimental results. Finally

the accurate prediction of the temperature rise of the GIS conductor is realized. Under the transient time-varying current

the maximum relative error of the constructed model is reduced from 26.28% to 1.31% compared with the MRT-LBM calculation results without data assimilation. The model is used to simulate the heat transfer inside GIS under different Ra conditions. The results show that with the increase of Ra

the heat transfer performance of GIS increases and the temperature rise decreases. The proposed method provides an explicit algorithm for accurately predicting the temperature rise of GIS under complex boundary constraints and exploring the internal heat transfer characteristics of GIS under different Ra.