本文提出了一种带有未知参数和输出噪声协方差的线性多变量系统的次最优控制算法。这种算法基于把辨识、估计和控制的函数明确分开的想法。以自举方式用随机近似方法估计系统的参数和状态。然后利用来自参数和状态估计器的估计导出确定的控制增益,得到次最优控制器。当估计的系统参数接近实际值时这个次最优控制器将接近于最优对策。 This paper presents a suboptimal control algorithm for a linear multivariable system with unknown parameters and output noise covariance. This algorithm is based on the idea of ​​explicitly separating the functions of identification, estimation and control. The system parameters and states are estimated by using the stochastic approximation method. The estimated control gain is then derived using the estimates from the parameter and state estimator to obtain the sub-optimal controller. When the estimated system parameters close to the actual value of this suboptimal controller will be close to the optimal solution.