Building Control Fault Prediction Using Operational State Coefficients
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing building automatic control systems face challenges in accurately predicting failures of facilities due to insufficient data for training artificial intelligence models, leading to high data processing requirements and costs, which are not feasible with limited-capacity systems.
Innovation Solution
An intelligent building automatic control system that predicts failures using operational state coefficients calculated from integrated values of electric power, static and dynamic pressure, and velocity head, with low data processing requirements, employing machine learning algorithms to enhance self-learning failure prediction.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If deep learning algorithms are used for failure prediction, then prediction accuracy is improved, but data processing requirements and system costs increase significantly
Solution Approach 1:
The patent extracts only the essential features needed for failure prediction from operational data, rather than processing complete raw datasets. By identifying and extracting key predictive features, the system achieves accurate failure prediction without requiring the substantial processing capacity needed for deep learning algorithms that process entire datasets.
Solution Approach 2:
Instead of reducing prediction accuracy to lower processing requirements, the patent inverts the approach by maintaining high prediction accuracy while reducing processing requirements through selective feature processing. This inverts the traditional trade-off by showing that accuracy can be preserved even with limited processing capacity.
2Measurement precision
If more operational data is collected for training AI models, then failure prediction accuracy is improved, but storage requirements and processing time increase
Solution Approach 1:
The patent extracts only the essential features needed for failure prediction from operational data, rather than processing complete raw datasets. By identifying and extracting key predictive features, the system achieves accurate failure prediction without requiring substantial processing time.
Solution Approach 2:
The patent applies partial action by processing only the necessary portion of operational data required for failure prediction. Instead of analyzing all available operational data, the system selectively processes relevant features, thereby reducing processing time while maintaining prediction accuracy.
3Reliability
If comprehensive facility monitoring is implemented, then failure detection capability is improved, but system complexity and costs increase
Solution Approach 1:
The patent extracts only the essential features needed for failure prediction from operational data, rather than processing complete raw datasets. By identifying and extracting key predictive features, the system achieves accurate failure prediction without requiring substantial processing time.
Solution Approach 2:
The system performs self-diagnosis and self-monitoring by automatically analyzing its own operational parameters. The facility monitoring system uses its own collected data to predict failures, eliminating the need for external complex analysis systems and reducing overall system complexity.
Data Source
AI summary
The present invention provides a building automatic control system capable of efficiently predicting failures of equipment installed inside a building using a limited-capacity computer system embedded within the building automatic control system.