A method and system for identifying frequent defect areas of a multi-cavity injection mold

By collecting and analyzing cavity structure and surface image data of multi-cavity injection molds, and combining deep learning and computer vision technologies, a dynamic correlation between defects and cavity structure is established. This solves the problem of inaccurate localization of internal defects in existing technologies, and improves the production quality and stability of new energy vehicle parts.

CN122415500APending Publication Date: 2026-07-17QINGDAO HAISHIHAO PLASTIC CO LTD
View PDF 0 Cites 0 Cited by

Patent Information

Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
QINGDAO HAISHIHAO PLASTIC CO LTD
Filing Date
2026-04-21
Publication Date
2026-07-17

AI Technical Summary

Technical Problem

Existing methods for identifying defects in multi-cavity injection molds cannot accurately locate the root cause area of ​​defects inside the mold, leading to misjudgment and omission of frequently occurring areas, thus reducing the accuracy of identifying frequently occurring defect areas in multi-cavity injection molds.

Method used

By collecting cavity structure data of multi-cavity injection molds and surface image data of injection molded parts, and combining deep learning algorithms and computer vision technology, defect features and location information are extracted, a dynamic correlation between defects and cavity structure is established, the occurrence pattern of defects is explored, and areas with frequent defects are identified.

Benefits of technology

It has achieved high-precision defect positioning of injection molded parts for battery casings and motor casings of new energy vehicles, avoiding misjudgment and omission, improving the production stability and pass rate of molds, and providing a scientific basis for mold maintenance and optimization.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN122415500A_ABST
    Figure CN122415500A_ABST
Patent Text Reader

Abstract

The present application relates to the technical field of defect area identification, and particularly relates to a multi-cavity injection mold defect frequent area identification method and system. The method comprises the following steps: acquiring cavity structure data of a multi-cavity injection mold and injection part surface image data; performing defect detection on the injection part surface image data to extract defect features and defect position information; combining the cavity structure data and the defect position information, locating a defect associated area corresponding to a mold cavity, and establishing a dynamic correlation between the defect position and the cavity structure parameters; acquiring time series data of defect occurrence, and mining defect occurrence rules of the defect associated area, including time distribution characteristics, frequency change trends and interval fluctuation rules of defect occurrence, to identify defect frequent candidate areas; and based on defect features, performing feature verification and positioning on the defect frequent candidate areas to output corresponding defect frequent areas. The present application can improve the identification accuracy of mold defect frequent areas.
Need to check novelty before this filing date? Find Prior Art