为联合系统辨识和控制器设计，在子空间辨识的基础上研究了一种新的子空间预测控制. 该控制方法可自动校正模型预测控制中的系统参数，避免传统的线性二次高斯最优控制中繁琐的设计过程，且不依赖于控制器的任何先验信息. 在带有约束条件时，利用椭球优化来迭代产生一系列体积逐渐减小的椭球序列，该序列最终收敛到一个最优解. 在此基础上推导了椭球优化算法达到收敛时所需迭代次数的一个上界. 以直升机悬停状态为例，利用该文方法设计控制器，验证了子空间预测控制方法的有效性.

Subspace predictive control based on subspace identification is discussed for joint system identification and control design. This combination enables automatic tuning of the parameters in model predictive control and avoids many steps in the LQG-controller design. It is independent of any priori information of the controller. Considering the constrain conditions, an ellipsoid optimization algorithm is proposed to generate a sequence of ellipsoids with decreasing volume. An upper bound of the maximum number of possible iteration steps is derived. As an example of helicopter, efficiency of the proposed subspace predictive control strategy is realized.