Machine Learning Fault Detection for Connected Building Equipment
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Building management systems (BMS) face challenges in accurately predicting and diagnosing faults in equipment, leading to potential equipment failures and increased maintenance costs, as existing methods rely on rule-based systems that are cumbersome and inefficient in adapting to changing conditions.
Innovation Solution
A predictive diagnostics system using supervised and unsupervised machine learning techniques to analyze temporal data from connected equipment, generating a probability distribution to separate normal and faulty conditions, and creating a fault prediction model to anticipate and diagnose faults before they occur, enabling proactive maintenance.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Adaptability or versatility
If rule-based systems are used for fault prediction, then the system structure is simple, but the adaptability to changing conditions deteriorates
Solution Approach 1:
The patent replaces traditional rule-based mechanical diagnostic systems with machine learning-based predictive diagnostics. The system uses supervised and unsupervised learning algorithms to automatically adapt to changing equipment conditions without requiring manual rule updates, thereby improving adaptability while managing complexity through automated model training.
Solution Approach 2:
The system dynamically adjusts diagnostic parameters and thresholds based on learned patterns from historical data. By continuously updating probability distributions and fault detection parameters through machine learning, the system adapts to changing conditions while maintaining a structured approach to fault prediction.
2Measurement precision
If machine learning techniques are implemented, then fault prediction accuracy is improved, but computational complexity increases
Solution Approach 1:
The patent segments the fault prediction process into distinct phases: data collection, probability distribution generation, fault detection, and diagnosis. By dividing the complex machine learning task into manageable segments with specific functions, the system achieves high accuracy while controlling computational complexity through modular processing.
Solution Approach 2:
The system performs preliminary data processing and probability distribution generation before actual fault detection. By pre-processing data and establishing baseline distributions in advance, the system reduces real-time computational requirements while maintaining high prediction accuracy during operational monitoring.
3Reliability
If continuous monitoring is performed, then fault detection capability is improved, but energy consumption increases
Solution Approach 1:
The patent implements periodic sampling of equipment data rather than continuous monitoring. By collecting data at optimized intervals and using machine learning to detect trends between samples, the system maintains high fault detection capability while significantly reducing energy consumption compared to continuous real-time monitoring.
Data Source
AI summary
A building management system includes connected equipment configured to measure a plurality of monitored variables and a predictive diagnostics system configured to receive the monitored variables from the connected equipment; generate a probability distribution of the plurality of monitored variables; determine a boundary for the probability distribution using a supervised machine learning technique to separate normal conditions from faulty conditions indicated by the plurality of monitored variables; separate the faulty conditions into sub-patterns using an unsupervised machine learning technique to generate a fault prediction model, each sub-pattern corresponding with a fault, and each fault associated with a fault diagnosis; receive a current set of the monitored variables from the connected equipment; determine whether the current set of monitored variables correspond with one of the sub-patterns of the fault prediction model to facilitate predicting whether a corresponding fault will occur; and determining the fault diagnosis associated with the predicted fault.