Computational Theoretical Chemistry TRIZ Cases
- ML Force Field Training With Off-Equilibrium Data Sampling
High-uncertainty off-equilibrium sampling improves ML force field accuracy across temperature and pressure ranges without full ab initio cost.
- Quantum Chemical Structure Optimization With Dual Convergence Criteria
Dual force and energy-rate convergence stops DFT structure optimization earlier, cutting computation time while preserving stable-structure accuracy.
- Organic Intermolecular Coupling Prediction with Graph Neural Networks
Graph neural networks replace repeated DFT calculations with rapid coupling estimates for organic molecules and charge-carrier mobility screening.
- Discrete-Surface Continuum Solvent Modeling for Faster Simulations
Discrete surface elements and field points reduce the cost of modeling solvent polarization, forces, and solvation energy around molecules.
- Blockwise Recursive PLS for Delayed Model Updates
Blockwise recursive PLS detects output changes and delays model replacement until validation shows better performance.
- Distributed Molecular Dynamics Using Local Coordinate Frames
Local coordinate frames enable distributed molecular dynamics computation across multiple processing nodes.
- Density Functional Theory Structure Relaxation Convergence Optimization
Information processing program executes structure relaxation calculations with optimized convergence conditions for electron density updates.
- Shale Oil Quantification via Kerogen-Graphene Simulation
Composite kerogen-graphene simulation models resolve adsorption accuracy limitations of simple carbon materials, enabling precise shale oil quantification.
- Molecular Force Field Multi-Objective Fitting Algorithm Library
Unified framework automating multi-objective optimization for force field parameters, resolving complexity from disparate software dependencies.