Grouping Power Sources and Loads by Proximity
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Managing complex systems with numerous power sources and loads becomes increasingly difficult due to the need to track dynamic connections and physical locations, especially in large data centers, where existing methods like Power Line Identification require excessive time and resources.
Innovation Solution
Divide power sources and loads into groups based on shared properties, such as relative physical proximity, to facilitate efficient management and matching, reducing computational effort and improving topology representation.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If Power Line Identification method is applied to large data centers with many servers and outlets, then complete topology information can be obtained, but the time and computing resources required increase quadratically
Solution Approach 1:
The patent divides the large system into multiple groups based on spatial proximity or organizational criteria. By segmenting the overall topology problem into smaller sub-problems that can be solved independently within each group, the computational complexity is reduced from quadratic O(n²) to linear or near-linear O(n), while still achieving complete and accurate topology information through the aggregation of group-level results.
2Measurement precision
If direct matching of all servers to all outlets is performed, then accurate power connections are identified, but computing resources increase nonlinearly
Solution Approach 1:
The patent segments the matching process by first grouping servers and outlets based on shared properties such as physical location, rack position, or power distribution zone. This creates smaller matching sub-problems within each group, reducing the number of comparisons needed from n×m to much smaller group-sized comparisons, thereby improving productivity while maintaining matching accuracy through property-based filtering.
Solution Approach 2:
The patent applies local quality by using property-based characteristics (such as physical proximity, rack location, or power zone) to filter and pre-select candidate matches. This local filtering approach reduces the search space for each matching operation, improving efficiency while ensuring that only relevant candidates are considered, thus maintaining accuracy without requiring exhaustive comparisons of all possible pairs.
3Reliability
If complete tracking of all power connections is maintained, then system reliability is improved, but system complexity increases
Solution Approach 1:
The patent segments the power management system into hierarchical levels (e.g., data center level, rack level, device level) or functional groups. This segmentation allows reliability tracking to be maintained through aggregated metrics at higher levels while detailed connection tracking is performed only within smaller groups, reducing overall management complexity while preserving system-wide reliability information.
Data Source
AI summary
A method of managing a system having a multitude of power sources and power loads, configured for execution in a computing device, the computing device being assigned to the system; and system having a multitude of power sources and power loads, wherein such a managing method is applied to the system.