Microgrid Load Controller Pattern Matching
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing micro-grid systems face high initial construction costs and lack flexibility when new electric power loads are added, as they rely on past load history data, making it difficult to adapt to changes in configuration.
Innovation Solution
A micro-grid system with a controller that measures load data for newly added loads, compares it to preset load patterns, and adjusts distributed power sources based on the comparison to optimize power generation and consumption.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Adaptability or versatility
If a micro-grid system is constructed with traditional load prediction methods using past load history data, then the system can operate with established loads, but it cannot flexibly adapt when new electric power loads are added to the system
Solution Approach 1:
The system segments the load prediction function by creating separate prediction modules for different load types. Each load type has its own prediction model trained on historical data specific to that load category, allowing independent adaptation when new loads of known types are added without reconfiguring the entire system.
Solution Approach 2:
The system performs preliminary classification of new loads into predefined load types based on their characteristics before applying prediction. This preliminary action enables the system to immediately use existing prediction models for the classified load type, avoiding the need for immediate system reconfiguration or new model training.
2Productivity
If distributed power sources are adjusted to maximize operating profits based on accurate load prediction, then system efficiency improves, but the requirement for accurate load prediction data increases system complexity
Solution Approach 1:
The system creates simplified copies of load patterns by identifying representative load types and using their historical prediction data as templates. When a new load is added, the system copies the prediction approach from the matching load type template, providing accurate predictions without building complex new prediction models from scratch.
Solution Approach 2:
The system changes parameters by transitioning from generic overall load prediction to type-specific prediction parameters. By categorizing loads into distinct types with characteristic patterns, the prediction model uses appropriate parameters for each type, improving accuracy while maintaining manageable system complexity through parameter specialization.
Data Source
AI summary
The present invention relates to a micro-grid system including a load. The load includes: a distributed power source including one or more elements which generate power; and a controller configured to derive a first load pattern by measuring load data of the load when the load is added to the micro-grid system established in advance, compare the first load pattern with a preset load pattern for each load type, and operate the distributed power source based on a result of the comparison.