Generation of large-bandwidth x-ray free electron laser with evolutionary many-objective optimization algorithm

Abstract

X-ray free-electron lasers (XFELs) are cutting-edge scientific instruments for a wide range of disciplines. Conventionally, the narrow bandwidth is pursued in an XFEL. However, in recent years, the large-bandwidth XFEL operation schemes are proposed for X-ray spectroscopy and Xray crystallography, in which over-compression is a promising scheme to produce broad-bandwidth XFEL pulses through increasing the electron beam energy chirp. In this paper, combining with the beam yaw correction to overcome the transverse slice misalignment caused by the coherent synchrotron radiation, finding out the over-compression working point of the linac is treated as a many-objective (having four or more objectives) optimization problem, thus the non-dominated sorting genetic algorithm III is applied to the beam dynamic optimization for the first time. Startto-end simulations demonstrate a full bandwidth of 4.6% for Shanghai soft x-ray free-electron laser user facility.

Combining multi-objective genetic algorithm and neural network dynamically for the complex optimization problems in physics

“…Not only can all kinds of RF cavity optimizations be well handled, such as multi-cell cavity 11 and heavy ion cavities 41 , but also other accelerator optimization are principally suitable, like free electron laser 5 , 6 , nonlinear beam dynamics 4 and so on. If we look at optimization designs in other physics fields, such as radio apparatus 42 and structural components of materials 43 , more questions would be solved.…”

“…The field of accelerator can be taken as an example. There are many complex optimization problems that do not have the known optimal solution sets (the Pareto front) in accelerator fields, and good performance has been preliminarily obtained with MOGAs in some of these problems, such as the optimization of lattice 4 , Free Electron Laser 5 , 6 and other accelerator facilities 7 . Designing the shape of radio frequency (RF) cavity is a sort of important problem as well, in which several objectives such as geometric shunt impedance (R/Q), Q factor and shunt impedance (R a ) are optimized simultaneously by tuning the geometric parameters of the cavity.…”

Incorporating Preference Information Interactively in NSGA-III by the Adaptation of Reference Vectors

“…A priori methods: In the article where NSGA-III was proposed [10], the authors suggested a mechanism to incorporate preferences a priori by filtering the RVs utilized by the method. A similar approach was used by Yan et al [34] and da Silva et al [8,31,30]. Cheng et al [6] proposed an a priori version of NSGA-III that requires the DM to specify a central RV, which is utilized to adapt the complete set of RVs.…”

Inhibition of Current-spike Formation Based on Longitudinal Phase Space Manipulation for High-Repetition-Rate X-ray FEL

“…With the rapid development of computer science and high-performance cluster, many improved algorithms are accessible and the Non-dominated Sorting Genetic Algorithm III(NSGA-III) is selected to be applied to this beam dynamics optimization strategy. NSGA-III is an evolutionary many-objective optimization algorithm using the reference-point-based non-dominated sorting approach, which keeps it efficient in optimizing more than three objectives simultaneously compared with NSGA-II [42,43,44]. Based on it, the beam dynamics simulation of the injector section is conducted using ASTRA, which is a space charge tracking algorithm widely used worldwide to simulate beam dynamics in injector sections [45].…”