Battery Generative Design Using ML-Driven Simulation Loops
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Battery manufacturing is a complex, iterative process that requires optimization across multiple parameters such as state of charge, chemistry, efficiency, safety, aging, temperature, durability, and sustainability, with existing methods lacking efficiency and effectiveness in generating optimal battery designs.
Innovation Solution
A computer-implemented method employing generative design using machine learning to automatically build models of energy storage devices, perform simulations, and evolve design parameters to achieve product and model design objectives, incorporating advanced data science, cheminformatics, and structure-based modeling to optimize chemical, configuration, and process spaces.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If traditional battery manufacturing methods are used, then manufacturing experience and existing processes are leveraged, but design efficiency and exploration of optimal solutions are limited
Solution Approach 1:
The patent creates virtual copies of battery designs through generative design software that generates hundreds or thousands of design options based on specified parameters. These digital models allow exploration of design space without physical prototyping, significantly improving design efficiency while managing complexity through simulation rather than physical experimentation.
Solution Approach 2:
The patent performs preliminary design exploration and optimization through generative algorithms before actual manufacturing. By pre-generating and evaluating numerous design options virtually, the system identifies optimal designs that meet multiple criteria (capacity, safety, cost, manufacturability) before committing to physical production, thus improving efficiency while managing manufacturing complexity.
2Adaptability or versatility
If generative design generates hundreds or thousands of design options, then design exploration and innovation are expanded, but computational resources and time required increase
Solution Approach 1:
The patent generates hundreds or thousands of design options beyond what a single designer could conceive, using generative algorithms to explore the full design space. This excessive generation of options ensures comprehensive exploration of possibilities, with subsequent filtering and selection processes identifying the most promising designs, thereby achieving high adaptability while managing computation time through automated evaluation.
Solution Approach 2:
The patent incorporates simulation and evaluation feedback loops that assess generated designs against specified criteria (manufacturability, performance, cost). This feedback mechanism allows the system to iteratively refine and select optimal designs from the generated options, enabling extensive design exploration while reducing computation time by focusing resources on promising candidates identified through automated evaluation.
3Ease of manufacture
If manufacturing constraints are incorporated into generative design, then manufacturability of designs is improved, but design flexibility and creativity are reduced
Solution Approach 1:
The patent specifies manufacturing constraints as parameters in the generative design process, allowing the system to generate designs that are both manufacturable and innovative. By defining constraints (e.g., manufacturing methods, material specifications) as input parameters, the system explores design space within feasible boundaries, maintaining design flexibility while ensuring manufacturability through parameter-driven generation rather than rigid rule-based filtering.
Data Source
Figure 1A
Figure 1B
Figure 1C
AI summary
A computer-implemented method and corresponding system perform generative design of an energy storage device. The method automatically builds at least one model of the energy storage device. The building is based on a design parameter space and employs a machine learning process. The method automatically performs a simulation of the energy storage device using the design parameter space, a design evaluation space, and the at least one model built. The performing produces at least one prediction. The method automatically evolves at least one of (i) the design parameter space and (ii) the design evaluation space. In an event the at least one prediction indicates that a product design objective or model design objective has been achieved, the method automatically converges on the design parameter space evolved, thereby completing a generative design of the energy storage device and, otherwise, repeats the building, performing, and evolving.