Single step phase optimisation for coherent beam combination using deep learning

Abstract

Coherent beam combination of multiple fibres can be used to overcome limitations such as the power handling capability of single fibre configurations. In such a scheme, the focal intensity profile is critically dependent upon the relative phase of each fibre and so precise control over the phase of each fibre channel is essential. Determining the required phase compensations from the focal intensity profile alone (as measured via a camera) is extremely challenging with a large number of fibres as the phase information is obfuscated. Whilst iterative methods exist for phase retrieval, in practice, due to phase noise within a fibre laser amplification system, a single step process with computational time on the scale of milliseconds is needed. Here, we show how a neural network can be used to identify the phases of each fibre from the focal intensity profile, in a single step of ~ 10 ms, for a simulated 3-ring hexagonal close-packed arrangement, containing 19 separate fibres and subsequently how this enables bespoke beam shaping. In addition, we show that deep learning can be used to determine whether a desired intensity profile is physically possible within the simulation. This, coupled with the demonstrated resilience against simulated experimental noise, indicates a strong potential for the application of deep learning for coherent beam combination.

Deep learning in airborne particulate matter sensing: a review

“…Future work could consist of using deep learning neural networks to design and optimise sensors. For example, deep neural networks have been used to optimise photonic crystal nanocavities [126], design and characterise plasmonic nanostructures [127], design chiral metamaterials [128] and enable the design of fibres for coherent beam combination [129]. As conceptualised in figure 10, in the future, one could ask an AI how to best design a sensor capable of low-cost sensing and identification of single particulate species.…”

The Self-Learning AI Controller for Adaptive Power Beaming With Fiber-Array Laser Transmitter System

“…There are numerous papers (see, for example, [11,12], and literature in these papers) dedicated to coherent beam combining using DNN with offline training strategy. Methods of DNN utilization for syntheses of optimal control are significantly varied for different authors and include direct generation of control by DNN using target-plane measurements [13], generation by DNN of some initial pupil-plane phase distributions [14], and more sophisticated cascaded schemes where DNN is operated in pair with SPGD [15]. However, this methodology has also several important drawbacks: (1) offline training technique supposes to manually collect or simulate huge training datasets that should cover all potential system application scenarios, (2) one should guarantee that optical system will always operate in the same environmental conditions as represented in this training set, so that any extension of condition ranges or any system modification automatically requires to supplement (in the best case) or completely rebuild (in the worst case) of training set with the following DNN retraining.…”

Optimisation of coherent beam combination using deep learning

“…The main goal of this work [10] was to develop a method for identifying the phase profile in a fiber array by only observing the intensity profile at or near the focus. The difficulty in identifying the phase arises because measuring the intensity profile only captures the intensity of the electric field, obscuring the phase information.…”