Research on kinematic calibration algorithm for a Stewart platform

In the optical system of a space camera, the Stewart platform is used for precise adjustment of the secondary mirror pose to compensate for aberrations and improve imaging quality. Due to machining and installation errors in the motion pairs of the Stewart platform, their pose adjustment accuracy is directly affected. To improve the motion accuracy of the Stewart platform, this paper proposes a Stewart platform kinematic calibration algorithm with the minimum residual of the driving amount. Firstly, 36 hinge point structural errors and 6 zero position strut length errors are randomly generated and added to the nominal structural parameters to simulate the actual mechanism parameters. Then, 12 sets of target poses are given, and 12 sets of 6-dimensional strut length arrays are obtained by inverse solving with nominal structural parameters. Then, 12 sets of strut length arrays are used to drive the mechanism model with errors, and 12 sets of poses are obtained through forward solution. This simulates the measured 12 sets of poses, and the difference between the measured pose and the target pose is the pose error before calibration. Then, the inverse kinematics model is applied, and the residual between the strut length is used as the optimization objective function. Nonlinear least squares technology is applied to identify the geometric parameter error of the motion platform. The simulation results show that the difference between the calculated actual structural parameters and the set actual structural parameters is within 0.1μm, within the allowable error range, which verifies the correctness of the kinematic calibration algorithm.