Flat wire motor card missing insertion online detection data processing system

By constructing a static spatial topology graph structure with slots as nodes and a graph convolutional network, and combining the timing signals of the insertion mechanism with the image stream, the visual features and timing data of the stator slots of the flat wire motor are fused, which solves the problem of missed insertion and misjudgment in dense occlusion scenarios and improves the accuracy of the detection system.

CN122391812APending Publication Date: 2026-07-14JIAXING SHILIAN INTELLIGENT TECH CO LTD
View PDF 0 Cites 0 Cited by

Patent Information

Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
JIAXING SHILIAN INTELLIGENT TECH CO LTD
Filing Date
2026-06-11
Publication Date
2026-07-14

AI Technical Summary

Technical Problem

Existing technologies cannot effectively solve the problem of missing insertion of the card coil in densely obstructed scenarios during the stator insertion process of flat wire motors, resulting in a high misjudgment rate and failing to utilize the physical space association of the slots with the timing logic of the insertion action.

Method used

A static spatial topology graph structure is constructed with slots as nodes and the physical distance between adjacent slots as edge weights. By combining the timing signals of the insertion mechanism and the image stream of the stator end, a graph convolutional network is used to perform joint iterative aggregation of visual features and temporal state labels to compensate for the visual features of occluded slots and dynamically adjust the edge weights to control the information attenuation of feature transmission.

Benefits of technology

It improves the recognition error tolerance and reliability of missing insertion state estimation under complex physical occlusion conditions, reduces the false judgment rate, and improves the accuracy of the detection system.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN122391812A_ABST
    Figure CN122391812A_ABST
Patent Text Reader

Abstract

The present application relates to the technical field of data processing, and discloses an online detection data processing system for flat wire motor hairpin missing insertion. A static space topology graph is constructed based on stator slot position distribution coordinates, taking slot positions as nodes and physical distances between adjacent slot positions as edge weights. Action timing signals of the wire insertion mechanism and stator end image streams are obtained, initial visual features of each slot end are extracted as node attributes, and action sequences in the action timing signals are converted into timing state labels to inject dynamic attributes of the nodes. In the graph convolution network, edge weights are taken as feature aggregation paths, and timing state labels and initial visual features of adjacent slot positions are iteratively aggregated, so that the initial visual features of the blocked slot positions are compensated by the visual features and timing state labels of the adjacent wire-inserted slot positions. The present application overcomes the misjudgment defect caused by the missing visual features under dense blocking, calculates the missing insertion state of the blocked area through topology neighborhood timing transmission, and improves the fault tolerance and reliability of detection.
Need to check novelty before this filing date? Find Prior Art