Design Optimisation/Simulation TRIZ Cases

Segmented sealing collar prevents drilling fluid leakage across the rock sample interface, ensuring accurate depleted formation simulation.

Replaces thin shell approximations with conical domain models to improve ionospheric slant TEC analysis accuracy.

A centralized social network tool integrates with video games to manage virtual member interactions and user profiles.

Enhanced behavioral models capture power-induced jitter effects in I/O buffers, resolving the trade-off between measurement precision and model complexity.

Dynamic model selection balances estimation accuracy against computational complexity, resolving inefficiencies in existing power consumption models.

Runtime incremental training refines hardness predictions using solved property results, improving scheduling accuracy and verification throughput.

A simulation system links historical plant data to virtual components for accurate emissions modeling.

A simulation method converts microring resonators into equivalent models to accelerate optical network-on-chip design verification.

Pre-refinement geometric map maintains consistency during mesh refinement, resolving convergence issues in isogeometric analysis.

Adjoint simulation adjusts structural parameters in physical devices to perform neural network inference, reducing computational cost and latency.

A neutron shield selection method calculates an overall figure of merit using absorption and scattering coefficients to identify optimal materials.

Federated learning trains placement models locally to reduce time and improve quality without sharing sensitive design data.

Screening variational neurons in neural networks builds transistor statistical models that capture process fluctuations without slow physics-based simulations.

Simplifies circuit layouts via rule-based modification to reduce electromagnetic simulation time while maintaining design fidelity.

Transformer models forecast stem cell differentiation outcomes to resolve protocol inefficiencies caused by donor variability.

A crop pest risk prediction system uses time-series environmental data to generate accurate disease forecasts.

Hybrid low-fidelity and high-fidelity modeling predicts CBR spread in complex buildings by adapting fidelity to computational constraints.

A recycling process planning engine generates optimized plans from manufacturing design data to maximize material recovery.

Dual grid system with information exchange unit resolves boundary processing complexity in lattice Boltzmann fluid analysis simulations.

A compound vector representation converts multi-dimensional CAD models into graph-based embeddings for machine learning classification.

Segmenting geometry into modules with varying grid densities reduces CAE analysis time while maintaining resin flow prediction accuracy.

Decomposing many-variable densities into few-variable signed densities enables efficient path integral Monte Carlo simulation on classical hardware.

A method predicts grease deterioration by measuring the initial storage modulus to loss modulus ratio and distortion at equal values.

Automated pattern flattening generates manufacturable 2D panels from 3D garment models for virtual draping.

Aspect-ratio scale factors modify Jacobian matrix derivatives in finite element hexahedral elements to reduce artificial stiffness.

A computer-based simulation method models bodily exudate distribution on absorbent substrates using finite element analysis.

A computer system maps chemical feature vectors to a two-dimensional space for intuitive user selection and structure generation.

Independent core simulations determine access order, enabling synchronized cache correction that resolves accuracy losses from race conditions.

A representative via selection method optimizes electrodeposition simulation using artificial neural networks.

A neural network model approximates physical quantity levels using spatial constraints represented by masking functions.

An autoencoder extracts large-scale features from observation data to guide an inversion network.

A multiscale CAE model simulates solder joint structural behavior during drop tests using global and local sub-models synchronized by kinematic constraints.

A graph convolutional network infers node displacements from stiffness matrices and force vectors.

Modal dynamic analysis uses a correction term to fix grossly incorrect stress and reaction force results from Lagrange multipliers.

Segmented impedance models predict substrate noise coupling to reduce computation speed while maintaining measurement precision.

A hybrid modeling system dynamically switches between a deep learning surrogate and a physics-based model to optimize computational performance.

A turbulent flow estimation method segments continuous fluid motion into discrete primitive shapes like polyhedrons and toroids for efficient modeling.

Electronic processor fits probit curves to test data, resolving Bruceton inaccuracies from non-step function distributions.

Optimizing an objective function with the L-BFGS algorithm maintains simulation stability for high elastic modulus models.

A computer-implemented methodology simulates plastic deformation in virtual threaded coupling models to derive accurate stress amplification factors.

A machine learning model predicts well corrosion severity ranks using barrier parameters and criticality features.

A semiconductor cooling system segments an integrated circuit to target heat dissipation based on regional reliability risk.

A method approximates chemical potentials in ternary and quaternary semiconductors using ab initio calculations.

A solver module uses a mesh-size independent stopping criterion to calculate potential solutions iteratively.

Precomputed influence coefficients replace iterative CFD calculations, reducing flow analysis time while maintaining measurement precision.

Multi-level feature selection reduces dimensionality of incineration process data to improve dioxin detection accuracy.

Clock skew scheduling optimizes integrated circuit timing by adjusting signal phases, preserving original topology to simplify debugging.

Optimized irregular sparse observation system designs shot and receiver points using forward modeling to reconstruct high-density seismic data.

A computer-implemented method uses volume-uniform basis functions to compute eddy currents in conductive bodies immersed in electromagnetic fields.

Particle motion equations with resilience terms simulate rubber contact states without calculation failures during large deformations.