Global Minimization of Longitudinal Emittance in RF Linac Structures

Relativistic electron bunches in linear accelerators are characterized by 6-D phase spaces. In most linear accelerators, the longitudinal phase space distribution does not interact significantly with the transverse distributions. This assumption allows the use of a 2-D design model of the longitudinal phase space. In this paper a global convergent parameter optimization is proposed to achieve an optimal bunch compression design with respect to an exact bunch energy, peak current and a vanishing skewness in the current profile, minimizing the longitudinal emittance by using a genetic algorithm and sequential quadratic programming hybrid formulation to find an optimal bunch.