To address the shortcomings of traditional motor thermal protection, which uses the motor’s temperature or temperature rise threshold operating at full load as the basis for the judgment method, a dynamic threshold judgment method based on a study of the motor’s temperature rise characteristics under a dynamic load rate is proposed. Taking a forced air-cooled, three-phase asynchronous motor with 11 kW power as the research object, a 2D/3D motor model under steady-state conditions is established. The loss distribution of asynchronous motors at different load rates is obtained through electromagnetic field simulation calculations. The finite element method is used to simulate and calculate the motor’s temperature field. Taking ambient temperature and load factors into account, the study takes place from the perspective of motor temperature rise in order to acquire the law governing the change in motor temperature increase under various working circumstances. A method of converting the temperature and warming between the motor’s stator, winding, and rotor is proposed. A dynamic threshold judgment system for accurate warning of overheating problems in motors under any load rate is set up so that the thermal protection of motors can accurately be judged under variable load conditions, which offers a crucial theoretical foundation for the design of intelligent motor temperature rise prediction systems. It contributes to a methodological underpinning for the diagnosis and early warning of thermal health in smart motor systems.