Building Equipment Predictive Control with Automated Horizon Selection
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing fault detection and prediction approaches in building management systems (BMS) rely on robust historical data with multiple instances of different fault types, which is often not available in practice.
Innovation Solution
A method for automatically selecting a prediction horizon for a predictive model used in BMS, involving evaluations of model performance at successively narrower ranges of possible prediction horizons until the optimal horizon is determined, allowing for automated control actions and fault predictions.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If traditional fault detection approaches are used that rely on robust historical data with multiple instances of different fault types, then prediction accuracy can be improved, but the system cannot operate effectively when such data is not available in practice
Solution Approach 1:
The system dynamically adjusts the prediction horizon parameter based on equipment-specific characteristics and available data, rather than relying on fixed historical fault patterns. This allows the predictive model to adapt to different equipment types and data availability scenarios, resolving the contradiction between maintaining prediction accuracy and achieving broad system applicability
Solution Approach 2:
The system performs preliminary evaluations of model performance at different prediction horizons before deploying the final predictive model. This preliminary tuning phase allows the system to optimize prediction accuracy for each specific equipment instance without requiring extensive historical fault data, thereby enabling effective operation across diverse equipment types
2Device complexity
If a fixed prediction horizon is used in the predictive model, then system complexity is reduced, but model performance cannot be optimized for different equipment and operating conditions
Solution Approach 1:
The system implements dynamic prediction horizon selection where the optimal horizon is automatically determined through performance evaluations at different time ranges. This dynamic approach allows the model to adapt to varying equipment behaviors and operating conditions while maintaining a relatively simple overall system architecture, resolving the contradiction between simplicity and reliability
3Measurement precision
If extensive historical data with multiple fault instances is collected and stored, then prediction model training can be improved, but data storage requirements and processing time increase
Solution Approach 1:
The system uses performance evaluations at selected prediction horizons rather than exhaustively analyzing all possible time ranges. This partial action approach achieves sufficient model training accuracy without requiring extensive data processing, thereby reducing the time loss while maintaining prediction effectiveness
Data Source
AI summary
A method includes automatically selecting a prediction horizon used by the predictive model by performing evaluations of model performance at successively narrower ranges of possible prediction horizons until the prediction horizon is determined based on results of the evaluations. The method may also include using the predictive model with the prediction horizon to perform an automated control action, which may include at least one of controlling or monitoring the building equipment.