Machine-Learning Threshold Adjustment for Building Air Quality Control
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current air quality detection systems in buildings take a one-size-fits-all approach, leading to inefficiencies and increased maintenance costs due to excessive wear and tear from limited response options, as they do not allow for adjustments or refinements in threshold values.
Innovation Solution
A building management system integrated with an air quality monitoring system that uses historical data to train machine-learning models to determine baseline and threshold values, enabling adjustments and mitigation actions for air quality management, thereby optimizing system responses.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If traditional air quality detection systems use fixed threshold values and limited response options, then the system structure remains simple, but the building system efficiency decreases and maintenance costs increase due to excessive wear and tear
Solution Approach 1:
The patent implements dynamic threshold values and response options that adapt based on historical air quality data and machine learning models. The building management system continuously learns from past air quality patterns and adjusts mitigation actions accordingly, transforming the static system into a dynamic one that optimizes building system efficiency while managing complexity through data-driven adaptation.
Solution Approach 2:
The system changes the parameter of threshold values from fixed to variable, using historical data to determine optimal thresholds. Machine learning models adjust these parameters dynamically based on learned patterns, allowing the system to improve efficiency by making informed decisions about when and how to activate building systems, thereby reducing unnecessary wear and tear.
2Ease of operation
If air quality detection systems allow for adjustments and refinements in threshold values, then the efficiency of building systems improves, but the device complexity increases
Solution Approach 1:
The building management system performs self-adjustment of threshold values and response parameters using machine learning models trained on historical air quality data. The system automatically learns optimal operating parameters without requiring manual intervention, improving ease of operation while managing complexity through automation rather than human expertise.
Solution Approach 2:
The system incorporates feedback loops where historical air quality data and system responses are continuously analyzed by machine learning models. This feedback mechanism allows the system to automatically refine threshold values and mitigation actions based on past performance, improving operability while the automation handles the complexity of continuous optimization.
3Reliability
If the building management system uses multiple response options based on historical air quality data, then maintenance costs decrease and system lifespan extends, but the complexity of decision-making increases
Solution Approach 1:
The system performs preliminary actions by pre-training machine learning models on historical air quality data to establish baseline threshold values and optimal response strategies. This preliminary learning phase enables the system to make reliable decisions in real-time without complex decision-making during operation, as the complexity has already been resolved through prior analysis of historical patterns.
Data Source
AI summary
An air quality monitoring system integrated with the building management system observes and records air quality data and building system data. Based on the observed air quality data and building system data, one or more training data sets can be generated. Such training data sets are used to train one or more machine-learning models configured to determine one or more threshold values for the air quality. These thresholds are then used to modify or adjust one or more building systems based on monitored air quality data.