Channel phase extraction for a coherent beam combining system using a 2D target intensity image and the CMA-ES algorithm

Abstract

We investigate a novel covariant matrix adaption-evolution strategy (CMA-ES)-based method proposed for extracting channel phase information by measuring a two-dimensional (2D) target intensity image (2D-TII) of a coherent beam combining (CBC) system both numerically and experimentally for the first time to the best of our knowledge. The proposed method was first investigated on 1,000 samples of 2D-TIIs numerically generated by a virtual 3-channel CBC system. For all samples, the channel phase information was extracted almost perfectly, with the inter-image correlation coefficient reaching or exceeding 0.99 and the overall root-mean-square phase error of 0.0735 rad within 17 iterations of the algorithm, for example. Next, the investigation was extended onto another 1,000 samples of 2D-TIIs experimentally measured with a real-world 3-channel CBC setup via a charge-coupled device (CCD) camera at a rate of 16 fps with an exposure time of 10 ms. The channel phase information was extracted with the inter-image correlation coefficient reaching or exceeding 0.9 for 972 or 979 samples within 15 or 45 iterations of the algorithm, respectively, with the latter case of which its overall average was estimated at 0.947. The relatively low performance of the proposed method within 21 out of 1,000 samples, where the overall average of the inter-image correlation coefficient remained at 0.880 regardless of further increases in the number of iterations, was attributed to the lowered image contrasts of the measured 2D-TIIs caused by the uncontrolled intrusion of external noise components that could not be rectified by the CCD camera due to its limited exposure time. We expect the proposed method to be useful for research and analysis on a variety of real-world CBC systems as well as other related applications where phase information needs to be extracted.