Abstract: In response to the problems of nonlinear uncertainty and time delay lag of the preheating temperature control in the selective laser sintering (SLS) process, a modified predictive fractional-order proportional-integral-differential (FOPID) algorithm for temperature control is proposed in this study. The designed controller firstly integrates the Smith predictor to eliminate the oscillations generated by the pure time lag chain, which is considered to improve the robustness of the controlled system. Then, a new novel sensitive parameter self-tuning method called EAOCBO is proposed to provide the optimal design of model parameters for the predictive FOPID controller. Aquila optimizer (AO) is a bionic intelligent algorithm, in order to overcome its shortcomings of the imbalance exploration and exploitation phase and ease to be trapped in the local optimum, the leader updating mechanism from COOT bird optimization, adaptive switching factor, and refracted opposition-based learning are introduced in the original algorithm. On the nine IEEE CEC2017 test functions, the optimization performance of EAOCBO is significantly enhanced compared to the other compared algorithms. In order to verify the effectiveness of the proposed EAOCBO-predictive FOPID controller, its dynamic response characteristics under unit step signal are simulated and analyzed by Matlab/Simulink software and further applied to the sintering machine to carry out the actual forming experiment. The results show that the proposed controller has lower regulation time and steady state error compared with other advanced FOPID controllers, which indicates its excellent response speed and control accuracy. Moreover, the EAOCBO-predictive FOPID controller can also precisely control the preheating temperature and improve the uniformity of the temperature field in practical applications, which in turn boosts the dimensional accuracy of the forming parts.