Design Support Device Using Machine Learning and Genetic Algorithms
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing design techniques face challenges in optimizing product design when the relationships between design factors and performance are unknown, leading to inefficient trial-and-error processes and suboptimal performance achievement.
Innovation Solution
A design support device and method that employs a machine learning model and genetic algorithm to estimate multiple performance types from design factors, generate Pareto solutions, and select optimal design factor groups through comprehensive evaluation and constraint-based optimization.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If trial production is performed to ascertain performance, then performance can be evaluated, but the process is time-consuming and requires repeated adjustments
Solution Approach 1:
The patent creates a virtual copy of the product through a machine learning model that predicts performance characteristics without physical trial production. The model learns from existing product data and generates performance estimates for new design configurations, eliminating the need for actual manufacturing and testing while maintaining evaluation accuracy.
Solution Approach 2:
The patent performs preliminary performance prediction using machine learning before actual trial production occurs. The system pre-evaluates multiple design configurations through computational modeling, allowing designers to identify promising candidates before committing to physical production, thus reducing the time required for iterative adjustments.
2Device complexity
If single regression analysis is used to determine design factors, then the analysis process is simple, but it cannot handle multiple performance types simultaneously
Solution Approach 1:
The patent employs a machine learning model that serves multiple functions simultaneously: it predicts various performance characteristics (strength, durability, cost, etc.) from design factors in a single integrated framework. This multi-functional model replaces multiple separate regression analyses, enabling comprehensive evaluation of multiple performance types without proportionally increasing complexity.
Solution Approach 2:
The patent combines multiple analysis approaches into a composite evaluation system that integrates machine learning predictions with design factor analysis. This composite approach synthesizes information from various performance dimensions into unified design recommendations, handling multiple performance types simultaneously while maintaining manageable complexity through integrated processing.
3Reliability
If design factors are adjusted to achieve one performance goal, then that performance improves, but other performance may decrease
Solution Approach 1:
The patent implements a dynamic evaluation system that continuously adjusts design factor recommendations based on the interrelationships between multiple performance goals. The machine learning model analyzes trade-offs in real-time and provides optimized design configurations that balance multiple performance requirements, allowing the system to adapt to different priority combinations of performance goals.
Solution Approach 2:
The patent utilizes parameter transformation and optimization techniques to convert multiple performance goals into a unified optimization framework. The system changes the representation of performance parameters through machine learning transformations, enabling simultaneous optimization of multiple objectives by finding design configurations where parameter changes in one area produce beneficial effects in other areas.
Data Source
AI summary
A design support device includes a processor configured to execute a program to estimate a plurality of types of performance of a product from a design factor group including a plurality of design factors of the product using a machine learning model, and generate candidates for design information of the product on the basis of comprehensive evaluation of the estimated plurality of types of performance. The processor is further configured to execute the program to acquire Pareto solutions to the plurality of types of performance as the candidates for design information using a genetic algorithm with the plurality of design factors as chromosome information.