Coke oven blowing cooler system is difficult to establish accurate mathematical model for its strong nonlinearity. To solve the problem, a predictive control strategy based on support vector regression(SVR) is proposed. SVR based on the structural risk minimization can directly reflect model nonlinear characteristics, and the adaptive weight particle swarm optimization(APSO) is utilized to optimize the SVR identification parameters. The rolling of the finite horizon optimization and the feedback correction of can predictive control which is the main body of the control system, overcome the uncertainty and nonlinear process effectively. On the MATLAB simulation platform, this control strategy is compared with the traditional PID(proportion integration differention). The simulation results show that the control strategy has strong anti-interference and robustness, which ensures the rapid and effective stability of the pre-cooling device in the process.