Based on the theory of viscoelasticity, a nonlinear dynamic model of the long-distance belt conveyor system is established considering PMSMs (Permanent Magnet Synchronous Motors). Then, a constraint following robust control is designed to accurately control the start and stop of the belt conveyor. The analysis then examines the speed, tension, and other parameters at critical positions during the startup phases with different desired speed curves. The parking phase is also investigated by dividing into four cases including unloaded free parking, fully loaded free parking, unloaded emergency parking, and fully loaded emergency parking. On this basis, four temperature conditions are set to study the slip ratio for the parking process: room temperature, extremely low temperature, monthly low temperature, and monthly high temperature. By comparing with traditional controllers, the proposed model-based underactuated controller can effectively improve the aforementioned metrics under various operating conditions, contributing to the safe and stable operation of long-distance nonlinear belt conveyors.