Alok Kanti Deb

A Method for Detecting Half-Broken Rotor Bar in Lightly Loaded Induction Motors Using Current

This paper presents an effective method of motor current signature analysis for detecting half- as well as full broken single rotor bar fault of a squirrel-cage induction machine under various loading conditions and speeds. The proposed method is based on spectral preprocessing of the stator current followed by subspace decomposition of the signal autocorrelation matrix to detect relatively low-amplitude fault sidebands. This method is found to be very effective in detecting low-amplitude sinusoids in a signal dominated by high-amplitude fundamental. The extended Kalman filter is used to estimate and track the fundamental component of the stator current. This component is subtracted from the measured stator current at every time step generating a resultant signal with a very low or negligible fundamental component. Subsequently, multiple-signal classification (MUSIC) is applied on the resultant stator current signal. Motor slip is estimated from principle slot harmonic to decide the approximate location of the fault sidebands. For effective fault detection, a hypothesis test is proposed to check the presence of sufficient fault frequency sideband in the current spectrum. This test works better if the lobe in the MUSIC plot due to the fault frequency is not distorted or overlapped by the fundamental component. Therefore, for each data window, the minimum size of the autocorrelation matrix is determined to generate distinct peaks. The proposed method applies to steady-state condition and is found to exhibit superior performance even during the light-load conditions with a half-broken bar.

MVEM-Based Fault Diagnosis of Automotive Engines Using Dempster–Shafer Theory and Multiple Hypotheses Testing

Internal combustion engines exhibit fast pulsating short-time dynamics due to the reciprocating cylinder motion, around mean operating points that change comparatively slow due to inputs such as throttle and load. Comparatively, simple mean value engine models (MVEM) describe the slow changes of the averaged states for automotive control and fault diagnosis. In this paper, a bank of state estimators based on MVEMs is used for fault residual generation. Three faults: 1) throttle mass air-flow sensor fault; 2) exhaust gas recirculation valve sensor fault; and 3) exhaust leak fault are considered here. These faults are significant as they affect emission levels. Optimized thresholds for residual classification are derived for minimizing false alarm rates and missed detection rates. The diagnosis logic, based on the principles of structured residuals proposed in literature, is extended here for multiple hypotheses testing. Furthermore, the Dempster-Shafer theory is used to associate a confidence measure with the decision conclusions and this is shown to improve isolation. Performance is demonstrated with automotive engine data obtained from a four-cylinder instantaneous spark-ignition engine (gasoline) system model, developed in the simulation software AMESim.

Determining Autocorrelation Matrix Size and Sampling Frequency for MUSIC Algorithm

Detectability of closely spaced sinusoids in a noisy signal using MUltiple SIgnal Classifier (MUSIC) depends to a great extent on the sampling frequency (F_s) and the size of the autocorrelation matrix (N). Improper choice of any of these may result in increased computational burden and/or unresolved frequency components. This paper presents an analytical approach to determine expressions of lobe width using F_s and N at lobe base (Δf_b) and half of the lobe height (Δf_h). The required values of F_s and N can be derived from the expression of Δf_b for distortion-less lobe heights of two closely spaced sinusoids. A tighter bound can be found using the expression of only Δf_h to resolve two distinct peaks. Probability of resolution using reciprocal of MUSIC peaks is determined for various N and its limit for full resolvability was verified with the derived analytical expressions.

Fault detection of Air Intake Systems of SI gasoline engines using mean value and within cycle models

This paper addresses the detection of faults in Air Intake Systems (AIS) of SI gasoline engines based on realtime measurements. It presents comparison of two classes of models for fault detection, namely those using a Mean Value EngineModel (MVEM) involving variables averaged over cycles andWithin Cycle Crank-angle-based Model (WCCM) involving instantaneous values of variables changing with crank angle. Numerical simulation results of intake manifold leak and mass air flow sensor gain faults, obtained using the industry standard software called AMESimTM, have been used to demonstrate the fault detection capabilities of individual approaches. Based on these results it is clear that the method using WCCM has a higher fault detection sensitivity compared to one that uses MVEM, albeit at the expense of increased computational and modeling complexity.

Development and validation of an MVEM from an SI-engine based WCCM

Mean Value Engine Models (MVEM) model the averaged dynamics of an automobile engine and have been used for the purpose of automotive control and fault diagnosis. Within-Cycle, Crank-Angle based Models (WCCM) are complex models that model the instantaneous dynamics of engine states and replace a real engine for simulation purposes. A WCCM was constructed on the simulation platform AMESim™. An MVEM was derived from this simulation data. The motivation for this work was to use this model in an MVEM based Fault Diagnoser

TS fuzzy model identification by a novel objective function based fuzzy clustering algorithm

A Fuzzy C Regression Model (FCRM) distance metric has been used in Competitive Agglomeration (CA) algorithm to obtain optimal number rules or construct optimal fuzzy subspaces in whole input output space. To construct fuzzy partition matrix in data space, a new objective function has been proposed that can handle geometrical shape of input data distribution and linear functional relationship between input and output feature space variable. Premise and consequence parameters of Takagi-Sugeno (TS) fuzzy model are also obtained from the proposed objective function. Linear coefficients of consequence part have been determined using the Weighted Recursive Least Square (WRLS) framework. Effectiveness of the proposed algorithm has been validated using a nonlinear benchmark model.

Adaptive neuro-fuzzy control of dynamical systems

In this paper, the an adaptive neuro-fuzzy control that combines the features of fuzzy sets and neural networks have been implemented and applied for the control of SISO and MIMO systems. Duffing forced oscillation system was considered as the SISO plant while the Twin Rotor laboratory set up that closely mimics helicopter dynamics was considered as the MIMO plant. The tracking performance of the controller has been demonstrated for time varying inputs. Robust performance of the controller was demonstrated by applying a pulse disturbance when the controlled plant had reached a steady state. Real time implementation of the controller has been demonstrated on the Twin Rotor system.

The need for bias modelling in MVEM based estimators

Mean Value Engine Models (MVEM) have been used extensively in automotive controls especially over the last 20 years. An MVEM was derived from a detailed Within-Cycle, Crank-Angle based Model (WCCM) that modelled the fluctuating cylinder combustion driven dynamics of a Spark Ignition engine. The model was designed for eventual use in a Fault Diagnoser built for an automobile engine system. While using this model in Extended Kalman Filter based estimators for fault residue generation, it was noted that the model suffered from biases that impaired the quality of estimation results. The biases were found to originate from the inherent simplifications associated with MVEMs. This led to an understanding of the limits of accuracy of a traditional MVEM model, the need for accurate bias modelling and the development of more robust estimators. Estimation results were found to improve after bias correction using Least-Square Support Vector Regressors.

Automated detection of newborn sleep apnea using video monitoring system

Automated detection of neonatal sleep apnea is essential for constrained environments with high patient to nurse ratio. Existing studies on apnea detection mostly target adults, and use invasive sensors. Few approaches detect apnea using video monitoring, by identifying absence of respiratory motion. They apply frame differencing and thresholding, not suitable for neonates due to their subtle respiratory motion intermixed with other body movements. Proposed method first applies motion magnification. Subsequently, it filters respiration motion using dynamic thresholding. The technique is benchmarked with simulated motion of varying respiration frequencies. When validated with neonatal video data, proposed method achieves both > 90% sensitivity and specificity.

Block Sparse Representations in Modified Fuzzy C-Regression Model Clustering Algorithm for TS Fuzzy Model Identification

A novel objective function based clustering algorithm has been introduced by considering linear functional relation between input-output data and geometrical shape of input data. Noisy data points are counted as a separate class and remaining good data points in the data set are considered as good clusters. This noise clustering concept has been taken into the proposed objective function to obtain the fuzzy partition matrix of product space data. Block orthogonal matching pursuit algorithm is applied to determine the optimal number of rules from the over specified number of rules (clusters). The obtained fuzzy partition matrix is used to determine the premise variable parameters of Takagi-Sugeno (TS) fuzzy model. Once, the premise variable parameters and optimal number of rules (clusters) are identified then formulate the rule construction for identification of linear coefficients of consequence parameters. The effectiveness of the proposed algorithm has been validated on two benchmark models.