Improved Iterative Quantum Algorithm for Ground‐State Preparation

Finding the ground state of a Hamiltonian system is great significance in many-body quantum physics and chemistry. We propose an improved iterative algorithm to prepare Hamiltonian. The crucial point optimize cost function on space via gradient descent (QGD) implemented devices. provide practical guideline selection learning rate QGD by finding fundamental upper bound establishing relationship between our first-order approximation imaginary time evolution. Furthermore, we adapt variational preparation method as subroutine generate ancillary utilizing only polylogarithmic resources. performance demonstrated numerical calculations deuteron molecule Heisenberg model without with noises. Compared existing algorithms, approach has advantages including higher success probability at each iteration, measurement precision-independent sampling complexity, lower gate resources are required when well prepared.