<p>Heat exchangers are the core components of energy transfer and conversion and are widely used in the energy, chemical, and other fields. In an actual operational process, load changes lead to variations in the operating conditions of the heat exchanger. Evaluating the heat-transfer performance is crucial for the safe and efficient operation of the system. To realize high-precision heat transfer prediction through simulations, instead of using traditional solid equipment, this study proposed a heat transfer prediction modeling method that combines three-dimensional high-precision and one-dimensional real-time dynamic simulations. This method combines the high-precision advantage of three-dimensional simulation with the real-time advantage of one-dimensional simulation. To verify the feasibility of the modeling method, a heat transfer prediction model was constructed based on the heat transfer channel structure of a CO₂ mixture heat transfer characteristic experimental test system. The steady-state and dynamic heat transfer characteristics of CO₂/R32 mixtures were simulated and experimentally tested. Finally, the real-time operational capability of the heat transfer prediction model was verified using a real-time simulator. The results showed that the heat transfer prediction model modeling method proposed in this study could improve the accuracy by 1.75–4.64 times compared with the conventional one-dimensional dynamic model. The established heat transfer prediction model exhibited good accuracy for both dynamic and steady-state processes. The average relative errors with the experimental results were in the range of 0.91%–2.83% under six sets of experimental tests. Thus, the proposed heat transfer prediction model can predict the heat transfer process in real-time under all experimental heat source conditions.</p>