The calciner is the key equipment in the new dry cement technology, the stable control of whose outlet temperature directly affects the quality of clinker. Focusing on the problems of large time delay and multiple disturbances in the calciner outlet temperature control, this paper proposes a feedforward-based generalized predictive control algorithm. First of all, by analyzing the mechanism model of calciner outlet temperature, the dominant factors and model structure are obtained. Secondly, a large number of real-time data is used to identify the autoregressive moving average with exogenous input (ARMAX) data-driven model. Finally, the feedforward-based generalized predictive control algorithm is proposed, and its simulation verification and semi-physical platform experiment are carried out. The results show that the proposed algorithm performs a significant control effect, reduces the overshoot, and shortens the response time.