在一类线性流形上讨论了来源于生产实践的矩阵方程(AX,XB)=(C,D)的最小二乘自反解,并且利用矩阵对的奇异值分解给出通解的一般表达式,同时解决了解对给定矩阵的惟一最佳逼近问题.
This paper discusses the least-squares solution of matrix equation(AX,XB)=(C,D) over reflexive matrices on Linear manifold.By applying the singular value decomposition,the expression of the general solutions and the unique optimal approximation solution to a given matrix are obtained.