针对有约束多目标多自由度预测控制问题 ,应用目标规划方法 ,提出了一种既适合于参数模型又适合于非参数模型的在线滚动优化策略 ,并且通过计算机仿真研究 ,验证了该方法的有效性 .然后 ,对于参数模型预测控制问题 ,提出了一种抗扰动的最小绝对值辨识算法 .由于该辨识算法可用目标规划快速求解 ,因此可作为慢时变工业过程控制的在线辨识算法 . Aiming at the constrained multi-objective and multi-degree-of-freedom predictive control problem, an objective of online rolling optimization strategy which is suitable for both parametric and nonparametric models is proposed by applying the target programming method. The computer simulation shows that the method is effective Then, for the problem of parameter model predictive control, a minimum anti-disturbance absolute value identification algorithm is proposed, which can be used as an online identification algorithm for industrial control of slow time-varying processes because it can be quickly solved by the target programming.