Abstract:In the design of a diffraction-limited storage ring (DLSR) consisting of compact multi-bend achromats (MBAs), it is challenging to simultaneously achieve an ultralow emittance and a satisfactory nonlinear performance, due to extremely large nonlinearities and limited tuning ranges of the element parameters. Nevertheless, taking the High Energy Photon Source (HEPS) as an example, we demonstrate that the potential of a DLSR design can be explored with a successive and iterative implementation of the particle swarm optimization (PSO) and multi-objective genetic algorithm (MOGA). It turns out that with a hybrid MBA lattice, it is feasible for the HEPS to attain a natural emittance of about 50 pm.rad, and meanwhile, realize a sufficient ring acceptance for beam accumulation with an on-axis longitudinal injection scheme. Particularly, this study indicates that a rational combination of the PSO and MOGA is more effective than either of them alone in approaching the true global optima, for an explorative multi-objective problem with many optimizing variables and local optima.