Complex anomaly for enhanced machine independent condition monitoring
- Authors: Amar, Muhammad , Gondal, Iqbal , Wilson, Campbell
- Date: 2015
- Type: Text , Conference proceedings
- Relation: 9th International Conference on Open Source Systems and Technologies, ICOSST 2015; Lahore, Pakistan; 17th-19th December 2015
- Full Text: false
- Description: Safety in machine applications requires tracking machine health during the time of operations. Anomaly detection techniques are used to model normal behavior of the machines and raise an alarm if any anomaly is observed. But traditional anomaly detection techniques do not identify type and severity of aberrance in terms of amplitude, pattern or both. Once the anomalous behavior is observed then fault detection techniques are applied to diagnose faults. For machine independent condition monitoring (MICM) a range of features transforms are needed for autonomous learning of the fault classifiers for different parameters to identify variety of fault types which requires huge amount of time. In this paper a novel complex anomaly plan (CAP) representation has been proposed with amplitude anomalies on real and pattern anomalies on imaginary axis. To plot amplitude and pattern anomalies in the CAP, normal state vibrations frequency features are used to train Gaussian models for each of the frequency. The dynamic location of the anomaly plotted in the CAP gives a measure of the intensity of the anomaly, where real and imaginary axis components help the fault classifier to make an appropriate selection of the transform and thus enhances the efficiency of MICM framework. © 2015 IEEE.
- Description: ICOSST 2015 - 2015 International Conference on Open Source Systems and Technologies, Proceedings
Vibration spectrum imaging : A novel bearing fault classification approach
- Authors: Amar, Muhammad , Gondal, Iqbal , Wilson, Campbell
- Date: 2015
- Type: Text , Journal article
- Relation: IEEE Transactions on Industrial Electronics Vol. 62, no. 1 (2015), p. 494-502
- Full Text: false
- Reviewed:
- Description: Incipient fault detection in low signal-to-noise ratio (SNR) conditions requires robust features for accurate condition-based machine health monitoring. Accurate fault classification is positively linked to the quality of features of the faults. Therefore, there is a need to enhance the quality of the features before classification. This paper presents a novel vibration spectrum imaging (VSI) feature enhancement procedure for low SNR conditions. An artificial neural network (ANN) has been used as a fault classifier using these enhanced features of the faults. The normalized amplitudes of spectral contents of the quasi-stationary time vibration signals are transformed into spectral images. A 2-D averaging filter and binary image conversion, with appropriate threshold selection, are used to filter and enhance the images for the training and testing of the ANN classifier. The proposed novel VSI augments and provides the visual representation of the characteristic vibration spectral features in an image form. This provides enhanced spectral images for ANN training and thus leads to a highly robust fault classifier.
Weighted ANN input layer for adaptive features selection for robust fault classification
- Authors: Amar, Muhammad , Gondal, Iqbal , Wilson, Campbell
- Date: 2015
- Type: Text , Conference proceedings
- Full Text: false
- Description: Model based feature selection for identification of diverse faults in rotary machines can significantly cost time and money and it is nearly impossible to model all faults under different operating environments. In this paper, feedforward ANN input-layer-weights have been used for the adaptive selection of the least number of features, without fault model information, reducing the computations significantly but assuring the required accuracy by mitigating the noise. In the proposed approach, under the assumption that presented features should be translation invariant, ANN uses entire set of spectral features from raw input vibration signal for training. Dominant features are then selected using input-layer-weights relative to a threshold value vector. Different instances of ANN are then trained and tested to calculate F1_score with the reduced dominant features at different SNRs for each threshold value. Trained ANN with best average classification accuracy among all ANN instances gives us required number of dominant features. © Springer International Publishing Switzerland 2015.
Autonomous behavior modeling approach for diverse anomaly detection application
- Authors: Amar, Muhammad , Wilson, Campbell , Gondal, Iqbal
- Date: 2014
- Type: Text , Conference paper
- Relation: ICOSST 2014 - 2014 International Conference on Open Source Systems and Technologies, Lahore, Pakistan, 18-20th Dec 2014 p. 122-127
- Full Text: false
- Reviewed:
- Description: For absolute process safety in diverse machine applications, timely and reliable anomalous behavior detection is very crucial. Different machine applications have different normal behavior patterns and safety standards thus require adjustable and adaptive anomaly detection techniques. In this paper an autonomous behavior modeling approach for anomaly detection has been presented. In this approach time segmented vibration signals from the machines are transformed into spectral contents. After normalization, these frequency domain contents are divided into weighted frequency bins and then Gaussian models are achieved for these frequency bins over the entire training set. Using summation rule on the outputs of Gaussian models a single indicative measure of the machine health: normality score is obtained. The sensitivity of the normality score and anomaly detector towards potential anomalous signals can be controlled by using different number of bins and weights. Suitable parameters values, number of bins and weights profile, for anomaly detector model are selected autonomously using minimum value of the cost function. The increase of normality score of this model above a certain threshold is considered an alarm indicating anomalous behavior. Thus the proposed method enables us to achieve autonomously a suitable anomaly detection model with suitable parameters with controlled sensitivity during the test phase.
Fuzzy logic inspired bearing fault-model membership estimation
- Authors: Amar, Muhammad , Gondal, Iqbal , Wilson, Campbell
- Date: 2013
- Type: Text , Conference paper
- Relation: 2013 IEEE Eighth International Conference on Intelligent Sensors, Sensor Networks and Information Processing p. 420-425
- Full Text: false
- Reviewed:
- Description: In rotary machines bearings are a primary cause of failure. In order to estimate the time before failure to provide information for timely bearing replacement strategies, condition-based machine health monitoring techniques are employed. This paper discusses a model for estimating the severity of bearing faults that can be used for residual bearing life estimation by processing the vibration signal. The proposed technique used in this model examines the spectral content of vibration signals across frequency bins and then fits Gaussian distributions to each frequency bin. With the use of these Gaussian models and training set examples with different fault severity levels, characteristic membership functions are constructed. This enables estimation of the severity levels of the bearing faults through a fuzzy-logic inspired process, whereby the severity level corresponds to the maximum of the set of corresponding membership functions. Thus based on discrete fault severity levels, trained Gaussian fittings of spectral bins and characteristic fault membership functions are capable to estimate the fault severity on a continuous scale.
Multi-size-window spectral augmentation: Neural network bearing fault classifier
- Authors: Amar, Muhammad , Gondal, Iqbal , Wilson, Campbell
- Date: 2013
- Type: Text , Conference paper
- Relation: 2013 8th IEEE Conference on Industrial Electronics and Applications (ICIEA) p. 261-266
- Full Text: false
- Reviewed:
- Description: Features extraction has always been crucial in rotary machines for Condition based machine health monitoring. Time-domain-segmentation being among the preliminary steps for further classification process plays a momentous role. Vibration signals from bearing are quasistationary in nature therefore calculation of constituent frequencies amplitudes in the vibration signal is dependent upon time-segmentation-window size. The proposed research confers the effects of time-segmentation window size on spectral features amplitudes calculation and its impacts on classification accuracy of the Artificial Neural Network (ANN). Using multi-size time-segmentation-window, for comprehensive spectral features calculation, ANN pattern classifier has been trained for enhanced classification. ANN learning assigns importance based relative weights to the links using supervised learning. Experimental results have shown that multi-size-window spectral features for ANN fault classifier perform efficiently for quasi-stationary bearing vibrations.
Unitary anomaly detection for ubiquitous safety in machine health monitoring
- Authors: Amar, Muhammad , Gondal, Iqbal , Wilson, Campbell
- Date: 2012
- Type: Text , Conference paper
- Relation: 19th International Conference on Neural Information Processing (INCONIP) p. 361-368
- Full Text: false
- Reviewed:
- Description: Safety has always been of vital concern in both industrial and home applications. Ensuring safety often requires certain quantifications regarding the inclusive behavior of the system under observation in order to determine deviations from normal behavior. In machine health monitoring, the vibration signal is of great importance for such measurements because it includes abundant information from several machine parts and surroundings that can influence machine behavior. This paper proposes a unitary anomaly detection technique (UAD) that, upon observation of abnormal behavior in the vibration signal, can trigger an alarm with an adjustable threshold in order to meet different safety requirements. The normalized amplitude of spectral contents of the quasi stationary time vibration signal are divided into frequency bins, and the summed amplitudes frequencies over bin are used as features. From a training set consisting of normal vibration signals, Gaussian distribution models are obtained for each feature, which are then used for anomaly detection.