Numerical Optimization Study Of Diffraction-Limited Storage Ring Lattices

At present,the fourth-generation synchrotron radiation sources,the diffraction-limited storage rings(DLSRs),are being developed rapidly around the world,which have beam emittances of hundreds of or even tens of pm·rad.In order to reduce the emittance,multi-bend achromat(MBA)lattices have been used in DLSR designs.Ultra-low emittance generally means that the lattice design will face very serious nonlinear beam dynamics effects.In order to improve the nonlinear dynamics performance,many DLSR lattice designs adopt very effective schemes of nonlinear dynamics cancellation.At present,numerical optimization methods based on intelligent optimization algorithms have been widely used in the lattice designs and optimizations.Compared with third-generation storage ring light sources,there are more optimization variables of magnet parameters and more complex and strict constraints in the DLSR lattice designs.For example,to cancel nonlinear dynamics effects,strict phase advance conditions need to be considered in the linear lattice design.When using an intelligent optimization algorithm to optimize the linear lattice of a DLSR,there is almost no solution satisfying the constraint conditions in the early stage of algorithm iteration,which will affect the optimization result.Therefore,in order to obtain better optimization results,we need to consider how to better deal with the constraint conditions.For a given linear lattice solution,the potential of its nonlinear dynamics optimization is limited.If we want to further improve the nonlinear dynamics performance of a DLSR,we need to explore the potential of linear lattice,which can be achieved by adding a guiding objective to represent the nonlinear dynamics to a certain extent when using the intelligent optimization algorithm to optimize linear lattice.This paper first focuses on the description of MBA lattices and intelligent optimization algorithms,including particle swarm optimization and genetic algorithm.Then in order to solve the problem of constraint processing in the numerical optimization of DSLR lattices,we take a hybrid 7BA lattice as an example,and optimize and compare several different constraint processing techniques with the multi-objective particle swarm optimization algorithm.To further improve the performance of nonlinear dynamics by exploring the potential of linear lattice,we also take this hybrid 7BA lattice as an example and construct guiding objectives which are characterized by linear parameters and optimized using multi-objective genetic algorithm.Then a comparison is made to illustrate the necessity and importance of using guiding objectives in the linear lattice design.The two numerical methods for lattice optimization based on the intelligent optimization algorithm studied in this paper will have certain reference significance and application value for DLSR lattice designs to obtain better linear lattice parameters and nonlinear dynamics performance.