Workpiece pose deviation is inevitable during automatic drilling and riveting in aircraft assembly, and it either leads to increased time cost or decreased position accuracy with the present methods. This study aims to propose a quick modification method to modify the coordinates of assembly holes based on online detection of the position of pre-assembly holes considering the effect of abnormal data caused by workpiece pose deviation. In the case of hole-making in the double curvature skin region of three pre-assembly holes, the space transformation matrix is determined using the unit quaternion method in accordance with the on-line detection of the hole position coordinates. In the case of more than three assembly holes with abnormal data, a new modification method using a random sampling consistency RANSAC algorithm is used to optimize the spatial transformation matrix and eliminate the abnormal data effectively. This method has been integrated into automatic drilling and riveting systems. Numerical simulation and experiment tests were conducted, and the result comparison showed that the method has obvious positive effect dealing with the abnormal data considering component pose deviation, which proves the effectiveness of the method.