Support Fault Diagnosis Research Articles

Data-Driven Fault Diagnosis in a Complex Hydraulic System based on Early Classification

In this paper, an early time-series classification (ETSC) algorithm is applied to support fault diagnosis in a complex hydraulic system (HS) with several interconnected components. The proposed technique aims at early classifying the state of the system while keeping the loss of classification inaccuracy at the minimum level. In contrast to baseline models that detect the eventual faults at the end of each working cycle, the ETSC model can diagnose any fault type of the HS components before observing the entire working cycle. Indeed, the early classification model successfully achieves a trade-off between the accuracy and the earliness criterion. Experimental results on a realistic HS dataset from the related literature show that the ETSC method can effectively identify different fault types with a higher accuracy and earlier compared to baseline methodologies.

Bearing Health Monitoring Using Relief-F-Based Feature Relevance Analysis and HMM

Nowadays, bearings installed in industrial electric motors are constituted as the primary mode of a failure affecting the global energy consumption. Since industries’ energy demand has a growing tendency, interest for efficient maintenance in electric motors is decisive. Vibration signals from bearings are employed commonly as a non-invasive approach to support fault diagnosis and severity evaluation of rotating machinery. However, vibration-based diagnosis poses a challenge concerning the signal properties, e.g., highly dynamic and non-stationary. Here, we introduce a knowledge-based tool to analyze multiple health conditions in bearings. Our approach includes a stochastic feature selection method, termed Stochastic Feature Selection (SFS), highlighting and interpreting relevant multi-domain attributes (time, frequency, and time–frequency) related to the bearing faults discriminability. In particular, a relief-F-based ranking and a Hidden Markov Model are trained under a windowing scheme to achieve our SFS. Obtained results in a public database demonstrate that our proposal is competitive compared to state-of-the-art algorithms concerning both the number of features selected and the classification accuracy.

Analytical Approach to Support Fault Diagnosis and Quality Control in End-Of-Line Testing

Operators in end-of-line testing of assembly lines often try out multiple solutions until they can solve a product quality issue. This calls for a decision support system based on data analytics that effectively helps operators in fault diagnosis and quality control. However, existing analytical approaches do not consider the specific data characteristics being prevalent in the area of End-of-Line (EoL) testing. We address this issue by proposing an analytical approach that is tailored to EoL testing. We show how to implement this approach in a real-world use case of a large automotive manufacturer, which reveals its potential to reduce unnecessary rework.

Case study on the effectiveness of gear fault diagnosis technique for gear tooth defects under fluctuating speed

Gears are very reliable parts of wind turbines and normally contribute for more than 5 years without failure when operated under fluctuating speed conditions. This case study presents the effectiveness of gear fault diagnosis techniques to highlight cracked tooth, chipped tooth and missing tooth gear under fluctuating speed conditions. Various measuring parameters such as overall vibration acceleration and envelope-detected acceleration were calculated for vibration signal, and sound intensity and sound pressure were calculated for acoustic signal. A statistical indicator kurtosis was also calculated for both vibration signal and acoustic signal. These measuring parameters were then compared with showing the fault detection capability of techniques employed under fluctuating speed and also under different loading conditions. The detection of faults by kurtosis of acoustic signal is found most efficient. For low load conditions, envelope-detected acceleration of vibration signal can also support fault diagnosis under fluctuating speed conditions.

A Wavelet Packet and Residual Analysis Based Method for Hydraulic Pump Health Diagnosis

This paper presents wavelet packet decomposition (WPD) and wavelet coefficient residual analysis based methods for hydraulic pump health diagnosis. A real-time pump health diagnosis system has been created on the basis of this method. This pump diagnosis system would analyse a short sequence of pump discharge pressure signals to detect if the pump was operating under a healthy condition or not. If the pump were operating with defective conditions, a further diagnosis would be implemented to identify the possible cause(s) of the defect(s). Based on the results obtained from a series of random laboratory tests by randomly selecting one of the four testing pumps 8 times, the developed WPD-residual analysis based pump diagnosis system was missing only one out of a total 32 diagnoses, which represented a 96.9 per cent accuracy rate in health diagnosis. Out of 23 fault diagnosing tests, 21 returned a correct diagnosis, resulting in a 91.3 per cent accuracy rate. The study also found that the accuracy rate could be further improved by taking available information from more packets to support fault diagnosis. It is worth pointing out that, while this method was tested against a hydraulic pump, it can also be applied to other equipment.

Architecture and Data Model for Monitoring of Distributed Automation Systems

A monitoring architecture and corresponding data model to support fault diagnosis in distributed automation systems is presented. The monitor system conSISts of a central monitor and a set of the monitoring agents, that cooperate via rule sets and diagnosis results. The data model combines event-based and continuous diagnostic methods and allows to adjust the monitor overhead. The achieved compacting of monitor information on the end nodes along with iterative diagnosis makes teleservice easier and more attractive.

Classification Method for Fault Diagnosis in Networked CIM Systems

A method to support fault diagnosis in networked discrete events systems is presented. The method combines the abstraction and condensation of data with recognition of anomalies. The method and implementing tool have been successfully used at CIM factories in some European countries.

A Monitor Tool for CIM Systems

Methods and an appropriated tool to support fault diagnosis in discrete events systems with network are presented. The methods combine the abstraction and condensation of data with recognition of anomalies. The tool has been successfully used at CIM factories in some European countries.

Civil aircraft maintenance and support Fault diagnosis from a business perspective

Aircraft maintenance down times together with maintenance activity durations and associated man‐hour expenditure are extremely important factors contributing to two major yardsticks of airline and civil aircraft performance: despatch reliability and direct maintenance costs, and have important cost implications. In maintaining an aircraft there is a need to predict fault diagnosis activities in quantitative terms of time. Traditionally estimating maintenance, in particular fault diagnosis, in terms of time has received scant attention. This paper identifies the need to evaluate all maintenance activity times. In addition, an approach is offered to estimate fault diagnosis activity times using knowledge‐based systems. Finally, the paper offers a vision outline in applying the technology and techniques to provide cost‐effective and timely fault diagnosis.