The prediction method for assembly poses deviation of satellite deployable arm considering force measurement uncertainty

Abstract Proper alignment between the rod flange and the joint flange of a satellite deployable arm is crucial for ensuring high assembly accuracy. Once these flanges are mated, the constrained positions generate assembly forces that cause rebound deformation of the arm, ultimately leading to pose deviation at the joint flange surfaces at both ends of the deployable arm. During assembly, multiple iterations of mounting, dismounting, and inspection are typically required, which is time-consuming and labor-intensive, thereby limiting assembly efficiency. To address this issue, a prediction method for assembly pose deviation in satellite deployable arms is proposed. By pre-evaluating the rebound pose deviation based on assembly forces, the method significantly reduces the need for repeated physical measurements and improves overall process efficiency. First, a finite element simulation model for assembly rebound is established. Subsequently, samples are obtained using Latin Hypercube Sampling (LHS). A stochastic Kriging model (SKM) is employed to construct a predictive model that relates assembly forces to pose deviation, taking force measurement uncertainty into account. Simulation experiments confirm the effectiveness of the proposed method, demonstrating that the average relative error remains below 1.7%. The high prediction accuracy verifies the feasibility and effectiveness of this approach.