A method and system for controlling CPU junction temperature
By processing temperature data through sliding median filtering and Z-score normalization, a temperature field grid topology map is constructed. By combining the spatiotemporal feature sequence and fused feature vector, the warning threshold and control parameters are dynamically adjusted, which solves the problem of inaccurate temperature control in the electricity consumption information collection terminal and realizes efficient temperature monitoring and control.
Patent Information
- Application Number
- CN202411706236.7
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- Filing Date
- 2024-11-26
- Publication Date
- 2025-10-10
- Estimated Expiration
- 2044-11-26
AI Technical Summary
Existing technologies lack collaborative analysis of multi-point temperature data in electricity consumption information collection terminals, resulting in inaccurate temperature field reconstruction, insufficient prediction of temperature anomaly trends, and fixed temperature thresholds that cannot adapt to environmental changes, resulting in inaccurate temperature control and high energy consumption.
Sliding median filtering and Z-score normalization are used to process temperature data, and a temperature field grid topology map is constructed. The spatiotemporal feature sequence and fusion feature vector are combined, and the temperature change is predicted through a gated recurrent unit network. The warning threshold and control parameters are dynamically adjusted to generate piecewise linear control instructions.
It achieves accurate identification and filtering of abnormal temperature data, improves the effectiveness and control accuracy of temperature monitoring, and improves the reliability and utilization efficiency of the terminal.