Logistics order distribution volume prediction method and device based on multi-level network point architecture
By employing sliding window anomaly detection, multi-level network hierarchy construction, multi-dimensional weight ranking, and user behavior learning optimization, combined with an integrated prediction model, the data quality and interactive experience issues in logistics delivery volume prediction were resolved, achieving efficient prediction and management.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- 上海乾臻信息科技有限公司
- Filing Date
- 2026-05-09
- Publication Date
- 2026-07-10
AI Technical Summary
Current logistics delivery volume forecasting suffers from several problems, including a lack of standardized data collection processes, untimely data updates, chaotic multi-level network hierarchy, a lack of scientific basis for network ranking, a forecasting model that does not integrate multi-dimensional influencing factors, and a poor user experience.
A sliding window anomaly detection model is used to identify and correct abnormal data, a multi-level network hierarchy is constructed, a multi-dimensional weight ranking model is introduced, user behavior learning is combined to optimize the display, an integrated prediction model is used to integrate multiple analysis methods, and a prediction error feedback mechanism is established.
It provides a high-quality data foundation, improves the convenience of branch management and the accuracy of predictions, enhances the interactive experience, reduces operating costs and improves service quality.
Smart Images

Figure CN122367320A_ABST