An official website of the United States government
Many industries are keenly interested in detecting and classifying faults before systems are sent to the customer or fail in use. A common approach is measuring the vibration of the machine and then using a classifier to check whether a fault is present. However, this process is difficult to automate because accelerometers are applied to the unit under test and are sometimes difficult to install and maintain due to complicated surface conditions. Accurate contact-based sensing is difficult when trying to check each rotating machinery assembly product during end-of-line quality control examinations or when evaluating the machine health of pre-installed rotating machinery. A deep learning-based fault classification system using both scalar and vector acoustic signals is a promising alternative that can replace the traditional error-prone, contact-based methods. Acoustic sound pressure and particle velocity measurements capture the directional fault signature of the mechanical defects in electric motors, and a one-dimensional convolutional neural networks (1D-CNNs) approach is proposed to process raw sensing data and eliminate the need for manual feature extraction. An experimental case study is performed to test the proposed 1D-CNN based fault classification on three different mechanically faulty electric motors across a variety of speeds. The results from acoustic pressure and particle velocity signals are compared against those from accelerometer signals. The experimental study confirms the feasibility of the proposed 1D-CNN on acoustic signals to be an excellent replacement for contact-based methods when assessing and classifying the machine fault condition. © 2025 Institute of Noise Control Engineering.
more »
« less
Deep learning based mechanical fault detection and diagnosis of electric motors using directional characteristics of acoustic signals
Early identification of rotating machinery faults is crucial to avoid catastrophic failures upon installation. Contact-based vibration acquisition approaches are traditionally used for the purpose of machine health monitoring and end-of-line quality control. In complex working conditions, it can be difficult to perform an accurate accelerometer based vibration test. Acoustic signals (sound pressure and particle velocity) also contain important information about the operating state of mechanical equipment and can be used to detect different faults. A deep learning approach, namely one-dimensional Convolution Neural Networks (1D-CNN) can directly process raw time signals thereby eliminating the human dependance on fault feature extraction. An experimental research study is conducted to test the proposed 1D-CNN methodology on three different electric motor faults. The results from the study indicate that the fault detection performance from the new acoustic-based method is very effective and thus can be a good replacement to the conventional accelerometer-based methods for detection and diagnosis of mechanical faults in electric motors.
more »
« less
- Award ID(s):
- 2015889
- PAR ID:
- 10428875
- Date Published:
- Journal Name:
- INTER-NOISE and NOISE-CON Congress and Conference Proceedings
- Volume:
- 266
- Issue:
- 2
- ISSN:
- 0736-2935
- Page Range / eLocation ID:
- 435 to 442
- Format(s):
- Medium: X
- Sponsoring Org:
- National Science Foundation
More Like this
-
-
Underground high-voltage transmission cables, especially high-pressure fluid-filled (HPFF) pipe-type cable systems, are critical components of urban power networks. These systems consist of insulated conductor cables housed within steel pipes filled with pressurized fluids that provide essential insulation and cooling. Despite their reliability, HPFF cables experience faults caused by insulation degradation, thermal expansion, and environmental stressors, which, due to their subtle and gradual nature, complicate incipient fault detection and subsequent fault localization. This study presents a novel, proactive, and retrofit-friendly predictive condition monitoring method. It leverages distributed accelerometer sensors non-intrusively mounted on the HPFF steel pipe within existing manholes to continuously monitor vibration signals in real time. A physics-enhanced convolutional neural network–long short-term memory (CNN–LSTM) deep learning architecture analyzes these signals to detect incipient faults before they evolve into critical failures. The CNN–LSTM model captures temporal dependencies in acoustic data streams, applying time-series analysis techniques tailored for the predictive condition monitoring of HPFF cables. Experimental validation uses vibration data from a scaled-down HPFF laboratory test setup, comparing normal operation to incipient fault events. The model reliably identifies subtle changes in sequential acoustic patterns indicative of incipient faults. Laboratory experimental results demonstrate a high accuracy of the physics-enhanced CNN–LSTM architecture for incipient fault detection with effective data feature extraction. This approach aims to support enhanced operational resilience and faster response times without intrusive infrastructure modifications, facilitating early intervention to mitigate service disruptions.more » « less
-
null (Ed.)Functional broadside tests were developed to avoid overtesting of delay faults. The tests achieve this goal by creating functional operation conditions during their functional capture cycles. To increase the achievable fault coverage, close-to-functional scan-based tests are allowed to deviate from functional operation conditions. This article suggests that a more comprehensive functional broadside test set can be obtained by replacing target faults that cannot be detected with faults that have similar (but not identical) detection conditions. A more comprehensive functional broadside test set has the advantage that it still maintains functional operation conditions. It covers the test holes created when target faults cannot be detected by detecting similar faults. The article considers the case where the target faults are transition faults. When a standard transition fault, with an extra delay of a single clock cycle, cannot be detected, an unspecified transition fault is used instead. An unspecified transition fault captures the behaviors of transition faults with different extra delays. When this fault cannot be detected, a stuck-at fault is used instead. A stuck-at fault has some of the detection conditions of a transition fault. Multicycle functional broadside tests are used to allow unspecified transition faults to be detected. As a by-product, test compaction also occurs. The structure of the test generation procedure accommodates the complexity of producing functional broadside tests by considering the target as well as replacement faults together. Experimental results for benchmark circuits demonstrate the fault coverage improvements achieved, and the effect on the number of tests.more » « less
-
BLDC motors are widely employed in various applications because of their high efficiency, reliability, and long operational life. For BLDC motors, any electric malfunctions in the commutation signal creation with or without Hall sensors can lead to unexpected vibrations, crashes, and accidents according to application areas. Therefore, fast detection and diagnosis of these faults are crucial for the reliable operation of BLDC motor drive systems. In this paper, a unique approach has been explored for developing fault signatures to detect commutation signal faults accurately and rapidly in the BLDC motor drive system under highly dynamic loads. After the fault detection, a commutation signal is indirectly reconstructed based on healthy commutation signals to continuously drive the motor drive system to avoid serious electrical and mechanical issues due to the faults. The proposed approach and feasibility of the method have been verified both by simulation and experimental studies. The results of the proposed method will significantly improve the accuracy of the commutation signal fault detection and eventually enhance the reliability of the BLDC motor drive.more » « less
-
A diagnostic test generation procedure targets fault pairs in a set of target faults with the goal of distinguishing all the fault pairs. When a fault pair cannot be distinguished, it prevents the diagnostic test set from providing information about the faults, and consequently, about defects whose diagnosis would have benefited from a diagnostic test for the indistinguishable fault pair. This is referred to in this paper as a diagnostic hole. The paper observes that it is possible to address diagnostic holes by targeting different but related fault pairs, possibly from a different fault model. As an example, the paper considers the case where diagnostic test generation is carried out for single stuck-at faults, and related bridging faults are used for addressing diagnostic holes. Considering fault detection, an undetectable single stuck-at fault implies that certain related bridging faults are undetectable. The paper observes that, even if a pair of single stuck-at faults is indistinguishable, a related pair of bridging faults may be distinguishable. Based on this observation, diagnostic tests for pairs of bridging faults are added to a diagnostic test set when the related single stuck-at faults are indistinguishable. Experimental results of defect diagnosis for defects that do not involve bridging faults demonstrate the importance of eliminating diagnostic holes.more » « less
- Free Publicly Accessible Full Text
-
Accepted Manuscript1.0
- Journal Article:
- https://doi.org/10.3397/NC_2023_0066
