Building Fault Detection Using CUSUM Fault Boundary Timing
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing building management systems struggle to accurately detect the beginning and end of faults, leading to inaccurate diagnosis and management of issues affecting energy consumption and other building operations.
Innovation Solution
A cumulative sum (CUSUM) analysis is performed on actual and predicted building data to identify faults by determining cumulative error values, gradients, and thresholds, allowing for precise identification of fault onset and cessation times.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If traditional fault detection methods are used, then the system can identify faults, but the boundaries (beginning and end) of faults cannot be accurately identified
Solution Approach 1:
The system performs preliminary actions by calculating cumulative sum values and identifying local minima before the fault is fully detected. By analyzing the cumulative sum trend in advance and detecting the local minimum point, the system can identify the fault start time earlier and more accurately, rather than waiting for traditional threshold-based detection methods.
Solution Approach 2:
The patent transforms the fault detection problem from a single-threshold comparison into a two-dimensional analysis by computing cumulative sum values over time and analyzing their trends. This dimensional transformation allows the system to identify fault boundaries by examining the shape and characteristics of the cumulative sum curve, including local minima and gradient changes, rather than relying on simple threshold crossings.
2Measurement precision
If cumulative sum analysis is performed to accurately identify fault boundaries, then fault timing precision improves, but computational complexity increases
Solution Approach 1:
The fault detection process is segmented into distinct computational stages: (1) calculating cumulative sum values from actual and predicted data, (2) identifying local minima points in the cumulative sum sequence, (3) computing gradients at these points, and (4) determining fault boundaries based on gradient thresholds. This segmentation allows the complex analysis to be broken down into manageable steps that can be efficiently implemented.
Solution Approach 2:
Instead of performing exhaustive analysis on all data points, the system applies partial action by focusing computational resources only on identifying local minima points and calculating gradients at these specific locations. This selective approach achieves accurate fault timing without the need to analyze every single data point in detail, thereby reducing overall computational complexity.
Data Source
AI summary
A building system for detecting faults in an operation of building equipment. The building system comprising one or more memory devices configured to store instructions thereon that cause one or more processors to perform a cumulative sum (CUSUM) analysis on actual building data and corresponding predicted building data to obtain cumulative sum values for a plurality of times within a first time period; determine a first time at which a first cumulative sum value is at a first maximum; identify a second cumulative sum value at a second maximum at a second time occurring after the first time; compare the identified second cumulative sum value to a threshold; and based on determining that the identified second cumulative sum value does not exceed the threshold, determine that a first fault ended at the first time.