Abstract: To address the challenges of significant friction prediction errors and high buckling risks during the running of completion string in ultra-long open-hole wells, a study on dynamic friction inversion and segmented control technology was conducted. Based on artificial intelligence clustering algorithms, the open-hole section is finely divided into low, medium, and high friction characteristic zones. Combined with the simulated annealing algorithm to dynamically correct the friction coefficient, a three-dimensional nonlinear mechanical model of the string-wellbore is established. Through researching the coupling effect of floating collar placement depth and drilling fluid density, a segmented regulation method based on friction zoning characteristics is proposed. This technology was applied in Well J108-2H. The optimized placement of the floating collars was determined at 2 600 m from bottomhole, and drilling fluids with densities of 1.42, 1.46, and 1.51 kg/L were used for regulation in the low, medium and high friction intervals respectively. This scheme controlled the equivalent circulating density within 1.90 kg/L, significantly reduced the buckling degree, and ensured the Φ139.7 mm casing string reached the target depth. The establishment of the dynamic friction inversion and segmented control method overcomes the nonlinear friction engineering limits brought by complex trajectories, providing a technical means for the safe extension of completion strings in ultra-long open-hole wells.