基于边缘计算网关的生产运行异常定位方法及系统
By eliminating prior production nodes through edge computing gateways and generating neural networks using adaptive and adversarial supervised learning, the accuracy problem of anomaly detection in industrial production is solved, enabling real-time monitoring and early warning, and improving production efficiency and quality.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- GUANGZHOU BOYITE INTELLIGENT INFORMATION TECH CO LTD
- Filing Date
- 2023-12-09
- Publication Date
- 2026-07-17
AI Technical Summary
Traditional data processing methods cannot accurately identify and locate anomalies in industrial production processes, especially when there is an imbalance of prior knowledge between different production nodes, which affects the accuracy of anomaly detection.
Sample production operation monitoring data is obtained through the edge computing gateway, prior production nodes are removed, a data sequence to be learned is generated, and an initial neural network is generated using adaptive knowledge learning and adversarial supervised knowledge learning. Verification and comparison are then performed to identify abnormal data, up to unlabeled non-prior nodes.
It enables real-time monitoring and early warning of industrial production processes, improving production efficiency and quality, and ensuring the accuracy of anomaly detection.
Smart Images
Figure CN117910541B_ABST