Design and fabrication of a robust hard X-ray multilayer using multi-objective genetic algorithms

A design approach based on multi-objective genetic algorithms (MOGAs) is proposed for the aperiodic multilayer design of an X-ray supermirror. Based on the nondominated solutions obtained through MOGA, the turning points between average reflectivity and flatness are identified. In the solution areas around these turning points, the design method based on MOGAs with an angle-based preference selection mechanism (MOGA-ANGLE) is employed to further explore the designed solutions, and an optimized solution has been obtained. Compared with the multilayer design method based on genetic algorithm, the experimental results demonstrate that the multilayer design obtained by our approach has higher flatness and stronger robustness, and the potential performances have been fully explored.