When machining complex curved parts on five-axis machine tools, due to factors such as the large curvature of the machining path and high feed acceleration, the tracking error of each axis cannot be controlled to be sufficiently small, resulting in difficulties in controlling the contouring accuracy. Coordinating the tracking error of each axis by matching servo gains emerges as a potential approach. However, existing gain solution methods are complex and time-consuming, rendering them inapplicable to actual production. In this paper, based on the explicit functional relationship between the contour error and the position loop gain (PLG) of each axis, a dynamic matching model for the PLG is established, and an intelligent control process for the linkage trajectory accuracy is proposed. Throug h “selfawareness” of the setpoint position and tracking error of each axis,“ self-analysis” of the position where the contour error of the linkage trajectory exceeds the limit, and“ self-decision” of the optimal contour error and the corresponding PLG, this intelligent control process can configure exclusive gains for different parts to ensure the machining accuracy of complex curved parts. For the KMC400 five-axis vertical machining center, the PLG of each axis is (70, 70, 80, 70, 75). After matching the exclusive gain (70, 68, 80, 59, 70) for the ruled surface A of the S–shaped test piece, the maximum contour error is reduced by 43.75%; after matching the exclusive gain (40, 70, 48, 40, 75) for the suction surface of the small blade of the impeller, the maximum contour error is reduced by 28.57%. The calculation time for the exclusive gains of the ruled surface A of the S–shaped test piece and the suction surface of the small blade of the impeller is 38.4 s and 5.7 s, respectively, meeting engineering application requirements.