The fine prediction of airflow distribution in aircraft cabins is an important basis for the study of cabin thermal comfort and biosafety. Due to the complex heat and mass transfer problems caused by forced convection and natural convection in large spaces, it is still faced with great challenges. In this paper, combined with the hybrid thermal lattice Boltzmann method (HTLBM) and GPU acceleration technology, a numerical simulation method of hybrid convection in the cabin suitable for engineering is developed, and a typical example is carried out to verify. Taking the cabin of a civil aircraft as the research object, the influence law of passenger behavior on the distribution of the flow field and temperature field in the cabin was compared and analyzed. The results show that large-scale circulation is significantly changed when passengers are standing in the aisle position, which increases the formation of the self-locking area of the passenger position and deflects the jet path of the air inlet, leading to heat accumulation near the aisle position.