Universal quantum state preparation via revised greedy algorithm

Preparation of quantum state lies at the heart information processing. The greedy algorithm provides a potential method to effectively prepare states. However, standard algorithm, in general, cannot take global maxima and instead becomes stuck on local maxima. Based this paper we propose revised version design dynamic pulses realize universal preparation, i.e., preparing any arbitrary from another one. As applications, implement scheme preparation single- two-qubit context semiconductor dots superconducting circuits. Evaluation results show that our outperforms alternative numerical optimizations with higher quality while possesses comparable high efficiency. Compared emerging machine learning, it shows better accessibility does not require training. Moreover, pulse sequences generated by are robust against various errors noises. Our opens new avenue optimization few-level system limited action space control problems.