一种基于物理残差约束与判别加权的刀具磨损预测方法

By constructing a tool wear prediction method based on physical residual constraints and discriminant weighting, the problem of insufficient generalization ability of existing tool wear prediction methods under insufficient samples and changing working conditions is solved, and accurate prediction of the entire tool wear life cycle is achieved, especially the key early warning in the accelerated failure stage.

CN122221703BActive Publication Date: 2026-07-17QILU INST OF TECH
View PDF 3 Cites 0 Cited by

Patent Information

Authority / Receiving Office
CN · China
Patent Type
Patents(China)
Current Assignee / Owner
QILU INST OF TECH
Filing Date
2026-05-19
Publication Date
2026-07-17

AI Technical Summary

Technical Problem

Existing tool wear prediction methods have insufficient generalization ability when there are insufficient samples, changes in working conditions, or noise interference, and they are difficult to fully characterize the three-stage evolution characteristics of tool wear throughout its entire life cycle, especially in the accelerated failure stage where learning is insufficient.

Method used

A tool wear prediction method based on physical residual constraints and discriminant weighting is constructed. By collecting multi-source feature data, a time coordinate variable and the three-stage evolution law of wear are constructed to form the physical residual of the wear prediction value. A joint loss function is constructed and a discriminant weighting coefficient is generated. The physical constraint weights of the samples are dynamically adjusted to optimize the model training.

Benefits of technology

It improves the physical consistency and training stability of the model, enhances the generalization ability under complex working conditions and the prediction accuracy of the critical stage of life, and provides a reliable basis for tool replacement and production scheduling.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN122221703B_ABST
    Figure CN122221703B_ABST
Patent Text Reader

Abstract

本申请公开了一种基于物理残差约束与判别加权的刀具磨损预测方法，涉及基于深度学习刀具磨损预测技术领域，依托刀具磨损三阶段演化规律构建物理残差，精准贴合全寿命周期磨损变化机理，弥补传统物理约束无法适配三阶段磨损特性的不足。同时构建数据与物理残差联合损失函数，依据物理残差难度生成判别加权系数，实现不同磨损阶段样本物理约束强度差异化调节。有效解决固定权重下模型对加速失效关键阶段学习不充分的问题，提升模型训练稳定性与物理一致性，弱化样本数量、工况变化及噪声干扰影响，增强复杂工况泛化能力与可解释性，精准捕捉寿命临界磨损变化，为刀具更换及生产调度提供可靠预测依据。
Need to check novelty before this filing date? Find Prior Art

Citation Information

Patent Citations

  • Digital twin cutter wear life prediction method and system based on physical information neural network

    CN119973726A

  • Deep learning rock freeze-thaw damage prediction method based on physical constraint

    CN120510951A

  • Single voltage prediction method of adaptive weighted physical information neural network

    CN121831541A