System and method for automatic detection and clustering of variable air volume units in a building management system
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional building management systems (BMS) are unable to automatically identify rogue, oversized, or undersized variable air volume (VAV) units, leading to energy wastage and inefficient energy management in HVAC systems, requiring manual and time-consuming processes for identification and correction.
Innovation Solution
A system and method for automatic detection and clustering of VAV units using machine learning models to tag suspect VAVs based on data analysis, including supply fan status, duct static pressure, and zone temperature, allowing for exclusion from duct static pressure setpoint calculations.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If manual identification and correction processes are used for rogue VAV units, then system reliability can be maintained through human expertise, but loss of time and productivity increase significantly
Solution Approach 1:
The system enables automatic self-diagnosis and identification of rogue VAV units through machine learning models that analyze operational data independently, eliminating the need for continuous manual monitoring and intervention while maintaining high reliability through automated detection capabilities
Solution Approach 2:
Manual mechanical processes of technician inspection and identification are replaced with automated computational systems using machine learning algorithms that process sensor data to identify rogue VAV units, dramatically reducing time loss while maintaining detection accuracy
2Reliability
If manual correction processes are used for rogue VAV units, then expertise can be applied to accurate diagnosis, but productivity decreases due to time-consuming procedures
Solution Approach 1:
The system automatically performs both identification and correction processes through automated algorithms that diagnose rogue VAV units and implement corrections without human intervention, maintaining diagnostic accuracy through sophisticated machine learning while dramatically increasing correction productivity
Solution Approach 2:
The machine learning models continuously analyze operational data in advance to identify potential rogue VAV units before they cause significant energy waste, enabling proactive corrections to be implemented automatically, thus maintaining high diagnostic accuracy while improving overall system productivity through preventive action
3Ease of operation
If all VAV units are included in duct static pressure setpoint calculations, then comprehensive control is achieved, but energy efficiency decreases due to rogue VAV influence
Solution Approach 1:
The system automatically extracts and excludes identified rogue VAV units from the duct static pressure setpoint calculations, separating problematic units from the overall control algorithm, thus eliminating their negative influence on energy efficiency while maintaining comprehensive control over legitimate VAV units
Solution Approach 2:
The system dynamically changes the parameter inclusion status in calculations based on real-time identification of rogue VAV units, automatically adjusting which units are included or excluded from pressure setpoint computations to optimize energy efficiency while maintaining appropriate comprehensive control
Data Source
AI summary
System and methods for automatic detection and clustering of variable air volume units in a building management system are disclosed. In one aspect, a method includes receiving one or more of variable air volume unit (VAV) data and an air handling unit (AHU) data from one or more data sources, determining one or more VAVs as a suspect VAV based on the VAV data and the AHU data, and removing the suspect VAV to determine a duct static pressure setpoint for an AHU.