As a universal framework for controlling nonlinear dynamic plants, the U-model shows significant simplicity and versatility advantages in controller system design. The goal of this study is to optimize the control system based on thermoelectric cooling, that is, apply the U-model method to the design of the thermoelectric cooling system, and compare it with the control effect of the battery thermal management system using model predictive control (MPC). Specifically, this study takes the battery thermal management model as an example and adopts two methods: controller design based on U-model and MPC control design. By comparing and analyzing the response time and control effect of the two control methods, the research results prove the superiority of the U-model in the design of an electric vehicle thermoelectric cooling control system. It has significant advantages in simplifying the design process and improving control performance. This comparative study further confirms that the U-model has broad application potential in optimizing the thermal management system of electric vehicles and provides strong support for designing future electric vehicle thermal management systems.