Use of Multiobjective Evolutionary Algorithms in High Brightness Electron Source Design

We describe the use of multiobjective evolutionary algorithms (MOEAs) for the design and optimization of a high average current, high brightness electron injector for an Energy Recovery Linac (ERL). By combining MOEAs with particle tracking, including space charge effects, and by employing parallel computing resources, we explored a multidimensional parameter space with 22 independent variables for a DC gun based injector which is being constructed at Cornell University. The simulated performance of the optimized injector is found to be excellent, with normalized rms emittances as low as 0.1 mm-mrad for a 77 pC bunch, and 0.7 mm-mrad for a 1 nC bunch. We detail the advantages and flexibility of MOEAs as a powerful tool well suited for wide application in solving various problems in the accelerator field.