Adaptive Quantum Algorithm Selection for Molecular Structure Optimization
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing quantum chemical calculations for molecule structure optimization face a trade-off between execution time and precision, with high-precision algorithms requiring excessive computational resources and low-precision algorithms lacking accuracy.
Innovation Solution
An algorithm selection program that dynamically switches between lightweight and high-precision quantum chemical calculation algorithms based on nuclear gradient norms, optimizing the number of cycles to minimize execution time while maintaining precision.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If a high-precision quantum chemical calculation algorithm is used for structure optimization, then the precision of the calculation result is improved, but the execution time increases excessively
Solution Approach 1:
The patent applies dynamics by making the algorithm selection adaptive and changeable during the calculation process. The system dynamically switches between different quantum chemical calculation algorithms based on the nuclear gradient norm at each cycle, transitioning from lightweight algorithms in early cycles to high-precision algorithms when convergence is approached, thereby optimizing both speed and accuracy throughout the optimization process
Solution Approach 2:
The patent changes the parameter of algorithm precision based on the calculation state. By monitoring the nuclear gradient norm and comparing it against thresholds, the system adjusts the algorithm precision parameter - using lower precision (lightweight algorithms) when the gradient is large and switching to higher precision (high-precision algorithms) when the gradient approaches the threshold, thus resolving the contradiction between execution time and calculation precision
2Productivity
If a lightweight quantum chemical calculation algorithm is used, then the execution time is reduced, but the precision of the calculation result deteriorates
Solution Approach 1:
The patent segments the structure optimization process into distinct phases based on the nuclear gradient norm. Early cycles with large gradients use lightweight algorithms for rapid progress, while later cycles near convergence switch to high-precision algorithms. This segmentation allows each phase to use the most appropriate algorithm precision, improving overall productivity without sacrificing final precision
Solution Approach 2:
The system dynamically adjusts algorithm precision to match the calculation phase. By monitoring the nuclear gradient norm, the system transitions from lightweight algorithms (higher speed, lower precision) in early cycles to high-precision algorithms (lower speed, higher precision) in later cycles, thereby achieving both fast execution and accurate results
Data Source
AI summary
A non-transitory computer-readable recording medium stores an algorithm selection program for causing a computer to execute a process including: in a case where a solution for structure optimization of a molecule is calculated using a quantum chemical calculation algorithm that repeatedly executes calculation for a plurality of cycles, executing a first quantum chemical calculation algorithm for a target molecule and calculating a nuclear gradient norm for each cycle; executing the first quantum chemical calculation algorithm for a cycle in which the nuclear gradient norm is equal to or more than a first threshold; and executing a second quantum chemical calculation algorithm of which execution time is longer than the first quantum chemical calculation algorithm for a cycle in which the nuclear gradient norm is less than the first threshold.