Dual matrices play an important role in robotics, kinematic analysis, spatial synthesis mechanisms, and so on. In this paper, two classes of inverse problems for dual matrices are discussed and the solvability conditions and the general solutions are achieved by applying the singular value decomposition and the generalized inverse. Further, the corresponding approximation problems are considered and the optimal approximate solutions are deduced by using some matrix decompositions and the Kronecker product. Finally, a numerical example is presented to verify the correctness of our results.
15A24, 65F18