Intelligent recommendation method for production process of non-oriented silicon steel
By constructing a predictive model based on knowledge graphs and machine learning, and combining it with a mechanistic model to optimize process parameters, the problems of long development cycles and high costs in traditional non-oriented silicon steel research and development have been solved, and rapid and accurate process parameter recommendations have been achieved.
CN121745253BActive Publication Date: 2026-06-05INST OF RES OF IRON & STEEL JIANGSU PROVINCE +2
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- INST OF RES OF IRON & STEEL JIANGSU PROVINCE
- Filing Date
- 2026-02-27
- Publication Date
- 2026-06-05
AI Technical Summary
Technical Problem
Traditional non-oriented silicon steel research and development relies on human experience and laboratory trial and error, resulting in long development cycles, high costs, and difficulty in quickly responding to market demands.
Method used
We construct a prediction model based on knowledge graphs and machine learning, combine mechanistic models and data-driven methods, and optimize process parameters through gradient descent to achieve intelligent recommendation.
Benefits of technology
It significantly shortens the product design cycle, reduces costs, improves the accuracy and reliability of process parameters, adapts to complex field environments, and enhances the model's generalization ability.
✦ Generated by Eureka AI based on patent content.
Smart Images
Figure CN121745253B_ABST
Abstract
The present application relates to a kind of intelligent recommendation method of non-oriented silicon steel production process, the present application is first based on the thickness of non-oriented silicon steel, component, process and performance standard data construction knowledge graph, after input target iron loss, through interpolation algorithm from knowledge graph obtains benchmark production path. Then, using industrial big data successively trains mechanism model and machine learning model, forms the prediction model of fusion mechanism and data. Subsequently, define iron loss optimization objective function, convert prediction model into product development model, and use gradient descent method to optimize and solve production path. Finally, the optimal production path obtained is put into industrial trial production, and the product development model is continuously optimized according to the feedback of trial production results, so as to continuously improve the prediction accuracy and practicability of product development model.
Need to check novelty before this filing date? Find Prior Art