Search All Patents in Quantum Computing
Patent US10452990
Issued 2019-10-22
Cost Function Deformation In Quantum Approximate Optimization
Techniques for performing cost function deformation in quantum approximate optimization are provided. The techniques include mapping a cost function associated with a combinatorial optimization problem to an optimization problem over allowed quantum states. A quantum Hamiltonian is constructed for the cost function, and a set of trial states are generated by a physical time evolution of the quantum hardware interspersed with control pulses. Aspects include measuring a quantum cost function for the trial states, determining a trial state resulting in optimal values, and deforming a Hamiltonian to find an optimal state and using the optimal state as a next starting state for a next optimization on a deformed Hamiltonian until an optimizer is determined with respect to a desired Hamiltonian.
Much More than Average Length Specification
View the Patent Matrix® Diagram to Explore the Claim Relationships
USPTO Full Text Publication >
- 1. A system, comprising:
a memory that stores computer executable components; a processor that executes computer executable components stored in the memory, wherein the computer executable components comprise:
a mapping component that maps a cost function to a Hamiltonian based on one or more constraints, wherein the Hamiltonian is associated with an optimization problem over allowed quantum states, and the Hamiltonian has a known answer at a time that the Hamiltonian is mapped;
a trial state and measurement component that generates trial states corresponding to the Hamiltonian by physical time evolution of quantum hardware interspersed with control pulses to entangle qubits of the quantum hardware, and that measures a quantum cost function for the trial states to determine a trial state that results in optimal values; and
a deformation component that deforms the Hamiltonian into a deformed Hamiltonian to find an optimal state, and uses the optimal state as a next starting state for a next optimization on the deformed Hamiltonian until an optimizer is determined with respect to a desired Hamiltonian.
- 5. A computer-implemented method, comprising:
facilitating, by a system operatively coupled to a processor, solving a combinatorial optimization problem using quantum hardware, the system comprising:
mapping, by the system, a cost function associated with the combinatorial optimization problem to an optimization problem over allowed quantum states, comprising constructing a quantum Hamiltonian for the cost function, wherein the quantum Hamiltonian has a known answer at a time of the construction;
generating, by the system, a set of trial states by a physical time evolution of the quantum hardware interspersed with control pulses;
measuring, by the system, a quantum cost function for the trial states;
determining, by the system, a trial state of the trial states resulting in optimal values; and
deforming, by the system, a Hamiltonian to find an optimal state and using the optimal state as a next starting state for a next optimization on a deformed Hamiltonian until an optimizer is determined with respect to a desired Hamiltonian.
- 13. A computer program product facilitating solving a combinatorial optimization problem, the computer program product comprising a computer readable storage medium having program instructions embodied therewith, the program instructions executable by a processor to cause the processor to:
map a cost function associated with the combinatorial optimization problem to an optimization problem over allowed quantum states, comprising constructing a quantum Hamiltonian for the cost function, wherein the quantum Hamiltonian has a known answer at a time of the construction; generate a set of trial states by a physical time evolution of the quantum hardware interspersed with control pulses; measure a quantum cost function for the trial states; determine a trial state resulting in optimal values; and deform a Hamiltonian to find an optimal state and using the optimal state as a next starting state for a next optimization on a deformed Hamiltonian until an optimizer is determined with respect to a desired Hamiltonian.
- 20. A computer-implemented method, comprising:
obtaining, by a system operatively coupled to a processor, a starting Hamiltonian having associated starting control parameters, wherein the starting Hamiltonian has a known answer at a time of the obtaining; using, by the system, quantum hardware to deform the starting Hamiltonian into a deformed Hamiltonian associated with optimal control parameters for that deformed Hamiltonian; using, by the system, the quantum hardware to repeatedly deform the deformed Hamiltonian with the associated optimal control parameters for that deformed Hamiltonian into further deformed Hamiltonians and further optimal control parameters associated therewith until a desired Hamiltonian is reached; and outputting, by the system, information corresponding to the desired Hamiltonian and the optimal control parameters associated with the desired Hamiltonian.
- 23. A computer program product facilitating solving a binary combinatorial optimization problem, the computer program product comprising a computer readable storage medium having program instructions embodied therewith, the program instructions executable by a processor to cause the processor to:
obtain a starting Hamiltonian having associated starting control parameters, wherein the starting Hamiltonian has a known answer at a time the starting Hamiltonian is obtained; use quantum hardware to deform the starting Hamiltonian into a deformed Hamiltonian associated with optimal control parameters for that deformed Hamiltonian; use the quantum hardware to repeatedly deform the deformed Hamiltonian with the associated optimal control parameters for the deformed Hamiltonian into further deformed Hamiltonians and further optimal control parameters associated therewith until a desired Hamiltonian is reached; and output information corresponding to the desired Hamiltonian and the optimal control parameters associated with the desired Hamiltonian.
