Building Vibration Anomaly Detection Using FFT and Machine Learning
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Analyzing large datasets from building equipment for operational issues is time-consuming and costly, especially with limited analysts, and requires manual parsing, which is inefficient and wasteful.
Innovation Solution
A building management system that uses machine learning models, specifically convolutional neural networks, to analyze vibration data sets by performing fast Fourier transforms and identifying abnormalities, reducing the workload for analysts by automating the analysis process.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If manual analysis by qualified analysts is used, then detection accuracy is maintained, but time consumption and cost increase significantly
Solution Approach 1:
The patent introduces machine learning models as an intermediary between raw vibration data and human analysts. The system automatically pre-processes and filters vibration datasets, generating preliminary analysis results that analysts can then review. This intermediary layer handles routine classification tasks, allowing human experts to focus on complex cases and thereby reducing overall analysis time while maintaining detection accuracy.
2Productivity
If more analysts are hired to handle large datasets, then analysis coverage improves, but training cost and operational complexity increase
Solution Approach 1:
The system enables self-service automation where machine learning models independently perform data classification, anomaly detection, and preliminary diagnosis without requiring human intervention for each dataset. The automated system serves itself by continuously processing vibration data, generating reports, and identifying patterns, thereby increasing analysis coverage without proportionally increasing the number of analysts or operational complexity.
3Measurement precision
If manual parsing of vibration data is performed, then detailed analysis is achieved, but efficiency decreases with large datasets
Solution Approach 1:
The patent segments the vibration data analysis process into distinct stages: automated preprocessing and feature extraction by machine learning models, followed by selective detailed analysis by human analysts for flagged anomalies. This segmentation allows the system to efficiently handle large datasets through automation while maintaining detailed analysis capabilities where needed, thereby improving overall productivity without sacrificing analysis detail.
Applied Scientific Principles
This section explains which scientific principles are used to turn an abstract innovation direction into a practical engineering solution.
Function Achieved in This Case
The system efficiently identifies normal and abnormal vibration data sets, reducing the burden on analysts by automatically flagging potential issues, ensuring no critical faults are missed and saving resources by minimizing unnecessary human analysis.
Implementation Method 1
The processing circuit is configured to perform one or more fast Fourier transforms on the vibration data set to generate a fast Fourier transform (FFT) spectra
Data Source
AI summary
A building management system including building equipment operable to affect a variable state or condition of a building. The building management system includes a controller including a processing circuit. The processing circuit is configured to obtain a vibration data set related to vibrations of the building equipment. The processing circuit is configured to analyze the vibration data set by one or more machine learning models to generate a set of probabilities. The set of probabilities is related to a probability that the vibration data set is abnormal. The processing circuit is configured to identify the vibration data set as normal or abnormal based on the set of probabilities. The processing circuit is configured to initiate a corrective action responsive to identifying the vibration data set as abnormal.