Closed Loop Simulation Platform for Polymer Electrolyte Discovery

Resolve Bottlenecks,
Find Innovative SolutionsGenerate Solutions

Solution Overview

Problem

The computational search for suitable polymer electrolyte materials for batteries is hindered by the high computational costs and time required for molecular dynamics simulations, making it impractical to screen a vast number of potential materials effectively.

Innovation Solution

A closed loop simulation platform that uses prediction models to rank and re-rank candidate systems based on predicted properties, autonomously reprioritizing simulations and reallocating resources, ensuring models stay updated with new data to accelerate the discovery of high-performance materials.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Measurement precision

If molecular dynamics simulations are performed to screen polymer electrolyte materials, then material properties can be accurately predicted, but computational cost and time consumption increase significantly

Engineering Contradiction:

Improveprediction accuracyVSAvoidcomputational time

Core Design Contradiction:

Measurement precisionVSLoss of time

Solution Approach 1:

The patent applies preliminary action by using machine learning prediction models to pre-screen and rank candidate materials before performing expensive molecular dynamics simulations. The system predicts material properties using compositional features and structural representations, identifies top-N candidates, and only then performs full simulations on these pre-selected materials, thereby avoiding unnecessary computational expenditure on unlikely candidates

Inventive Principle:

Principle #10Preliminary action

Solution Approach 2:

The patent introduces an intermediary machine learning model that acts as a mediator between the vast material space and the computationally expensive molecular dynamics simulations. This intermediary model, trained on compositional features and structural representations, filters and prioritizes candidates, enabling the simulation system to focus computational resources on the most promising materials while maintaining prediction accuracy

Inventive Principle:

Principle #24Intermediary (Mediator)

2Measurement precision

If molecular dynamics simulations are performed to screen polymer electrolyte materials, then material properties can be accurately predicted, but computational resources are consumed excessively

Engineering Contradiction:

Improveprediction accuracyVSAvoidcomputational resources

Core Design Contradiction:

Measurement precisionVSUse of energy by moving object

Solution Approach 1:

The system performs preliminary ranking using machine learning models that require minimal computational resources compared to full molecular dynamics simulations. By predicting material properties from compositional features and structural representations beforehand, the system identifies top-N candidates that warrant expensive simulations, thereby dramatically reducing overall computational resource consumption while preserving prediction accuracy for the selected candidates

Inventive Principle:

Principle #10Preliminary action

Solution Approach 2:

The patent creates simplified representations or copies of the full molecular dynamics simulation process through machine learning models. These surrogate models, trained on compositional features and structural data, approximate the behavior of expensive simulations, allowing the system to evaluate numerous candidates at low computational cost and only invoke full simulations on the most promising top-N candidates

Inventive Principle:

Principle #26Copying

3Reliability

If traditional computational screening methods are used, then all candidate materials can be evaluated, but the discovery process is too slow to be practical

Engineering Contradiction:

Improvecomprehensive evaluationVSAvoiddiscovery speed

Core Design Contradiction:

ReliabilityVSProductivity

Solution Approach 1:

The patent applies preliminary action by using machine learning models to rapidly evaluate and rank all candidate materials based on compositional features and structural representations before performing detailed molecular dynamics simulations. This pre-screening step maintains comprehensive evaluation by considering all candidates initially, then accelerates discovery by focusing expensive simulations only on top-N ranked candidates, achieving both reliability and productivity

Inventive Principle:

Principle #10Preliminary action

Solution Approach 2:

The system dynamically adapts its computational strategy by using machine learning models to identify promising candidates and adjust the allocation of computational resources accordingly. The top-N selection process is dynamic, allowing the system to focus computational power on the most promising materials at each stage, thereby accelerating the discovery process while maintaining thorough evaluation of the full candidate space through the predictive models

Inventive Principle:

Principle #15Dynamics

Data Source

PatentUS20240012964A1Closed loop simulation platform for accelerated polymer electrolyte material discovery

Publication Date: 2024.01.11 TOYOTA RESEARCH INSTITUTE INC

AI summary

A method of closed loop simulation for accelerated material discovery is described. The method includes ranking a plurality of candidate systems according to corresponding properties of interest predicted by a first prediction model. The method also includes simulating a first top-N of the plurality of candidate systems according to the corresponding properties of interest predicted by the first prediction model. The method further includes re-ranking the plurality of candidate systems according to the corresponding properties of interest predicted by a second prediction model. The method also includes simulating a second top-N of the plurality of candidate systems according to the corresponding properties of interest predicted by the second prediction model.