Behavior of the refrigerant flow within components that make up vapor compression refrigeration cycle, such as heat exchangers, compressors, and expansion valves, demonstrates inherent nonlinearity. Development controllers for air conditioning systems demand significant experimental time to control these nonlinear dynamics stably. In this study, we developed a controller to regulate both the indoor temperature and the discharge temperature of the compressor. The controller employs Model Predictive Control (MPC), which is adept at handling the complexities of nonlinear and MIMO systems. MPC has more control parameters than traditional methods like PID, making the tuning process time-consuming. Therefore, we introduced a model-based development (MBD) framework for designing air conditioner controllers that uses mathematical models and simulation environments. Due to the challenges associated with tuning control parameters experimentally using actual systems, utilizing simulation-based environments becomes invaluable for reducing the tuning workload.We constructed a model of the air conditioner using a nonlinear model. This allowed us to effectively replicate the behavior of the refrigeration cycle in simulations.We automated the tuning process of the MPC by employing a multi-objective optimization method, enabling us to optimize the control parameters.