一种基于AI训练与推理协同调度的指标口径冲突识别方法与系统
By using AI training and inference collaborative scheduling, an ordered transaction time list and a conflict identification model were constructed, which solved the problem of inventory data distortion caused by unclear ownership of goods in transit during multi-warehouse transfers, and achieved dynamic accuracy and real-time monitoring of inventory data.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- GUANGZHOU ZHONGYUAN INFORMATION TECH CO LTD
- Filing Date
- 2025-12-05
- Publication Date
- 2026-07-17
AI Technical Summary
During multi-warehouse transfers, unclear ownership of goods in transit leads to distorted inventory data. Existing technologies struggle to accurately identify and resolve inventory data discrepancies caused by overlapping transit statuses and transaction sequences.
By using AI training and inference collaborative scheduling, an ordered transaction time list is constructed, concurrent time series and time series overlap groups are identified, the degree of time series overlap is quantified, a caliber conflict identification model is constructed based on historical transfer records, inventory differences are analyzed, the ownership status of goods in transit is determined, a real inventory time series is generated, and a caliber conflict identification alarm is triggered.
It achieves dynamic accuracy of inventory data in complex allocation scenarios, automatically identifies and warns of inventory anomalies, and ensures real-time consistency and accuracy of inventory data.
Smart Images
Figure CN121920925B_ABST