Hybrid Quantum-Classical System for Multi-Objective Optimization
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current quantum computing technologies lack a native approach to efficiently solve multi-objective optimization problems, particularly on noisy intermediate-scale quantum (NISQ) devices, which are limited in size and prone to noise, and existing methods either rely on classical multi-objective optimization or require fully fault-tolerant quantum hardware.
Innovation Solution
A hybrid quantum-classical computer system utilizes quantum superposition to encode a population of solutions, with a parametrized quantum circuit generating a quantum state that represents multiple solutions simultaneously, allowing for the iterative selection and optimization of Pareto-optimal solutions, enabling the direct handling of multi-objective optimization problems on current and future quantum hardware.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Adaptability or versatility
If quantum approximate optimization algorithm (QAOA) or variational quantum eigensolver (VQE) is used to solve single objective optimization tasks, then the optimal solution can be found by minimizing a predefined cost function, but the method cannot handle multi-objective optimization problems with multiple conflicting objectives
Solution Approach 1:
The patent segments the multi-objective optimization problem into multiple single-objective subproblems by introducing penalty parameters. Each objective function is handled separately with its own penalty weight, allowing the quantum computer to solve each subproblem using existing single-objective algorithms like QAOA or VQE. This segmentation enables multi-objective capability while maintaining compatibility with current quantum hardware and algorithms.
Solution Approach 2:
The patent introduces penalty parameters as intermediary variables that mediate between multiple conflicting objectives and the quantum cost function. These penalty parameters act as intermediaries that translate multi-objective requirements into a form that single-objective quantum algorithms can process, bridging the gap between multi-objective problem formulation and single-objective quantum solution methods.
2Use of energy by moving object
If classical multi-objective optimization algorithms are used, then multi-objective problems can be solved, but the quantum aspect of quantum computation is not utilized and the solution quality may be limited by classical computational constraints
Solution Approach 1:
The patent merges classical penalty parameter optimization with quantum single-objective optimization algorithms. The classical component handles the multi-objective penalty parameter adjustment, while the quantum component solves the resulting single-objective subproblems. This hybrid approach combines the strengths of both classical and quantum computation, utilizing quantum resources to solve optimization subproblems while maintaining classical control over the multi-objective balancing.
3Ease of manufacture
If noisy intermediate-scale quantum (NISQ) devices are used, then quantum computation can be performed with current hardware limitations, but the devices are subject to noise and limited in size
Solution Approach 1:
The patent applies partial action by breaking down the multi-objective optimization into multiple simpler single-objective subproblems that can be solved sequentially or iteratively on NISQ devices. Rather than attempting to solve the full multi-objective problem in one complex quantum computation, the method performs multiple simpler quantum computations with different penalty parameter settings, accumulating results to approximate the Pareto front. This approach is feasible with current NISQ hardware limitations.
4Quantity of substance
If the target is to find a whole set of Pareto-optimal solutions rather than a single optimal solution, then all conflicting objectives can be addressed, but the computational complexity and number of required evaluations increase
Solution Approach 1:
The patent employs periodic action by iteratively solving multiple single-objective subproblems with different penalty parameter configurations. Each iteration focuses on optimizing a particular weighted combination of objectives, and by periodically varying the penalty parameters across different iterations, the method systematically explores different regions of the solution space to build up the complete Pareto front over time.
Data Source
AI summary
A computer implemented method for obtaining a set of Pareto-optimal solutions to a given multi-objective optimization problem with different objectives. The method generates a set of parameters on a classical computation component using the classical optimization algorithm. Based on the set of parameters, a quantum component generates a quantum state by executing a parameterized quantum circuit. A set of basis states is selected from the one quantum state The classical computation component calculates a set of objective values consisting of the objective values for each solution in the set of classical solutions, and a new estimate for the Pareto set of solutions and the corresponding Pareto front by the classical optimizer is generated. The internal state of the classical optimization algorithm is updated based on the current set of solutions and the Pareto set and their corresponding sets of objective values.