Abstract
High-dimensional optimization is a critical challenge for operating large-scale scientific facilities. We apply a physics-informed Gaussian process (GP) optimizer to tune a complex system. Typical GP models learn from past observations to make predictions, but this reduces their applicability to systems where there is limited relevant archive data. Instead, here we use a fast approximate model from physics simulations to design the GP model. The GP is then employed to make inferences from sequential online observations in order to optimize the system. Simulation and experimental studies were carried out to demonstrate the method for online control of a storage ring. Our method is a simple prescription to construct a custom GP model, including correlations between the high-dimensional input space, while encoding the physical response of a system. The ability to inform the machine-learning model with physics, without relying on the availability and range of prior data, may have wide applications in science.
Highlights
Online control and tuning of scientific experiments and facilities, such as free electron lasers and storage ring light sources, are challenging tasks, since those systems often consist of hundreds of correlated parameters that could be adjusted in order to find a set of parameter values to achieve optimal target performance
Automated tuning can help deliver the highest beam quality to scientific users during operation and reduce tuning time for operation mode switching. This would be enabled by efficient online optimization algorithms, which are necessary in particle accelerators because, physics models exist, there are often significant differences between the simulation and the real accelerator
We show that online optimization using the Gaussian process (GP) optimizers surpasses the current established optimization algorithms, Nelder-Mead simplex [1] and robust conjugate direction search (RCDS) [2], which are routinely used to tune particle accelerator systems [41]
Summary
Online control and tuning of scientific experiments and facilities, such as free electron lasers and storage ring light sources, are challenging tasks, since those systems often consist of hundreds of correlated parameters that could be adjusted in order to find a set of parameter values to achieve optimal target performance. Automated tuning can help deliver the highest beam quality to scientific users during operation and reduce tuning time for operation mode switching. This would be enabled by efficient online optimization algorithms, which are necessary in particle accelerators because, physics models exist, there are often significant differences between the simulation and the real accelerator. The critical requirement for a suitable tuning algorithm is the ability to find the optimum in a complex parameter space both robustly and with high efficiency (minimum number of steps). Traditional model-independent optimization methods, that do not require the gradient of the system, such as the standard optimizer Nelder-Mead simplex [1], may not work well for online applications when the target is noisy.
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have