Simulation Apparatus Using Macro-Model Learning to Reduce Calculation Burden
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Combining micro-simulation models for entire product simulations increases calculation complexity due to the need for finer time units and interaction simulations, which existing methods struggle to address effectively.
Innovation Solution
A simulation apparatus that performs coupled simulations using micro-models, learns the correspondence between inputs and outputs over time, and generates a macro-model to infer outputs after a period, reducing the need for detailed micro-model simulations in overall simulations.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If multiple micro-models are combined for coupled simulation to simulate entire product behavior, then simulation accuracy and comprehensiveness are improved, but calculation amount and device complexity increase significantly
Solution Approach 1:
The patent segments the simulation system into multiple independent micro-models, each representing a specific element or component. These micro-models can be developed, validated, and managed independently, then coupled together through standardized interfaces to simulate overall product behavior. This segmentation allows high simulation accuracy through detailed element-level modeling while managing complexity through modular architecture.
2Measurement precision
If micro-models use fine time units for detailed simulation, then simulation precision is improved, but calculation amount increases due to the need for frequent interaction simulations
Solution Approach 1:
The patent implements a periodic coupling mechanism where micro-models are simulated independently over coarse time intervals, and interactions are updated at periodic checkpoints. This allows each micro-model to maintain high precision with fine time steps internally, while the overall coupled simulation progresses efficiently through coarse time units. The periodic synchronization reduces the total number of interaction calculations required compared to continuous fine-grained coupling.
Data Source
AI summary
The present disclosure relates to a simulation apparatus and includes: a coupled simulation unit that performs coupled simulation to combine a plurality of micro-models each being a simulation model for a limited element; a macro-model learning unit that learns, from a result of the coupled simulation performed by the coupled simulation unit, a correspondence between an input and an output in overall simulation to simulate an entire simulation target and generates a macro-model to infer an output after a period of time has elapsed for an input; and an overall simulation performing unit that performs the overall simulation using the macro-model generated by the macro-model learning unit.