Enhancing pointing accuracy in Risley prisms through error calibration and stochastic parallel gradient descent inverse solution method

Abstract

High-precision pointing is crucial in evaluating the effectiveness of beam control technology. However, devices utilizing compact rotational Risley prisms (RRP) face a challenge in balancing field-of-view (FOV) and pointing accuracy. This study aims to develop a beam pointing model for RRP and enhance pointing accuracy through error calibration using the system identification method. Moreover, the current two-step inverse solution method is inadequate when the beam deflection model includes the prism's tilt error. To overcome this limitation, this study proposes a stochastic parallel gradient descent (SPGD) inverse solution method to further improve the system's pointing accuracy. A RRP system with a FOV of $\pm {15}^{\circ }$ was constructed for experimentation. The results indicate the pointing accuracy is 2.8 arcsec, the angular dynamic range is 46 dB, and the root-mean-square errors (RMSE) in the x- and y-directions are 1.2 arcsec and 0.9 arcsec, respectively. This device exhibits the highest angular dynamic range among existing RRP devices.