A current monitoring system for diagnosing electrical failures in induction motors
Introduction
Induction motors are essential components in the vast majority of industrial processes. The different faults on induction machines may yield drastic consequences for an industrial process. The main problems are related to increasing costs, and worsening of process safety conditions and final product quality. Many of these faults show themselves gradually. Then the detection of incipient faults allows avoiding unexpected factory stops and saving a great deal of money [1], [2]. The kind of faults of these machines are varied. However the most frequent are [3]:
- (a)
opening or shorting of one or more of a stator phase winding,
- (b)
broken rotor bar or cracked rotor end-rings,
- (c)
static or dynamic air-gap irregularities, and
- (d)
bearing failures.
These faults may be observed through some of the following symptoms [4]:
- (a)
unbalanced air-gap voltages and line currents,
- (b)
increased torque pulsations,
- (c)
decreased average torque,
- (d)
increased losses and
- (e)
excessive heating.
The reason for such faults may reside in small errors during motor manufacturing, improper use, high level of requirements in motor start-up, ventilation deficiency, and others. Motors actuated by pulse width modulation (PWM) voltage source inverters, have greater probabilities to fail in their bearings [5] and in their stator windings’ insulation [6].
Several diagnosis techniques for the identification and discrimination of the enumerated faults have been proposed. Temperature measurements, infrared recognition, radio frequency emissions, noise monitoring or chemical analysis are some of them [4]. References for coils to monitor the motor axial flux may be found in [7], vibration measurement, in [8], [9]. Spectrum analysis of machine line current (called motor current signature analysis or MCSA) is referred to in [10], [11], Park's Vector currents (PVC) Monitoring, in [12], [13], artificial intelligence based techniques are used in [14], [15], [16].
From all these approaches proposed in the literature, those based on stator current monitoring are advantageous because of its non-invasive feature. One of these techniques is the MCSA, in which rotor faults become apparent by harmonic components around the supply frequency. The amplitude of these lateral bands allows dimensioning the failure's degree [4]. Also, the Extended Park's Vector Approach (EPVA), based on the observation of the Park's complex vector module, allows the detection of inter-turn short circuits in the stator winding. This article presents the development of an on-line current monitoring system (CMS) to perform the diagnosis task in a supervisory system [17]. This last task employs both techniques (MCSA and EPVA) in an integrated way, for fault detection and diagnosis in the stator and in the rotor of an induction motor, respectively. The selection of both techniques, due to MCSA as well as EPVA, shares the stator current sensing, and then the same information may be used as input for both methods. In this way, current spectral components convey information about the rotor state, while the EPVA is appropriate for the stator windings monitoring, as it will be shown. The proposed CMS uses a National™ data acquisition equipment and is programmed in LabView™. From the acquired current data and the motor features, the CMS estimates the slip and load percentage. Based on experimental observations and on the knowledge of the electrical machine, a knowledge-based system (KBS) was constructed in order to carry out the diagnosis task from these estimated data. The results of each diagnosis are outcomes in the CMS screen in the form of fault modes index. If necessary, a warning is given to put the motor under new observations (i.e. to measure the rotor speed or to change the motor load), or even to verify the power distribution net balance. Experimental results are presented from an induction motor of 380 V, 7.5 HP and 1000 rpm, especially designed for running under different failure circumstances. These results with a high degree of correct diagnosis show a right direction to explore.
Section snippets
Motor current signature analysis
When there are broken or even fissured bars, the rotor's impedance exhibits an unbalance. The immediate consequence of such an unbalance is the existence of inverse sequence currents. These currents have a frequency that is equal to the product of the slip (s) and the supply frequency (f). They generate a magnetic field that turns counter motor rotation-wise. This magnetic field is called inverse magnetic field or IMF. The speed of this IMF is given by the expression (1):where ωri is
Towards an integrated fault diagnostic system
One aim of the present work was to combine the previous techniques and, in some way, to take the better of them in a single, integrated diagnostic system.
From the stator spectral analysis (MCSA) it is possible to detect rotor as well as stator winding faults, as presented in [20]. However, in the last case, the frequency characterising the fault must be computed considering the motor poles number, the slip, and the winding features. Also, another handicap of this approach is that it is not
The experimental prototype
The CMS proposed in this work consists of a data acquisition system sampling the stator currents from current transformers. The test bank is an induction motor dragging a direct current generator with variable load. The acquisition and human-machine interface tasks were developed with LabView. In Fig. 8 there is a schematic diagram, and in Fig. 9, a snapshot of the experimental prototype is used in this research.
The motor under diagnosis for these tests consists of an induction motor of 6 poles
Conclusions
Based on several techniques for fault detection and diagnosis of induction machines proposed in the literature, the ones considered most promising were selected. The selection criteria were: non-invasive technique, minimum number of measured variables, discrimination power, and prior motor information to yield a diagnosis.
The obtained CMS gives general conclusions about the motor state, in a user-friendly interface. It was easily developed in Lab with commercial products. The added feature of a
References (23)
- et al.
Basic tasks for knowledge based supervision in process control
Engineering Applications of Artificial Intelligence
(2001) - et al.
Reliability improvement and economic benefits of on-line monitoring system for large induction machines
IEEE Transactions on Industry Applications
(1990) - et al.
A survey of faults on induction motors in offshore oil industry, petrochemical industry, gas terminals, and oil refineries
IEEE Transactions on Industry Applications
(1995) - et al.
Cause and analysis of stator ans rotor failures in three-phase squirrel-cage induction motors
IEEE Transactions on Industry Applications
(1992) - S. Nandi, H. Toliyat, Fault diagnosis of electrical machines–a review. IEEE IEMCD’99, International Electric Machines...
- et al.
The effects of PWM voltage source inverter on the mechanical performence of rolling bearings
IEEE Transactions on Industry Applications
(1997) - et al.
Electrical stress and failure mechanism of the winding insulation in PWM-Inverter-Fed low-voltage induction motors
IEEE Transactions on Industrial Electronics
(2000) - et al.
Noninvasive detection of broken rotor bars in operating induction motors
IEEE Transactions on Energy Conversion
(1988) - et al.
Vibration and current monitoring for detecting airgap eccentricity in large induction motors
IEE Proceedings
(1986) - et al.
Analysis of airgap flux, current, and vibration signals as a function of the combination of static and dynamic airgap eccentricity in 3-phase induction motors
IEEE Transactions on Industry Applications
(1997)
An unsupervised, on-line system for induction motor fault detection using stator current monitoring
IEEE Transactions on Industry Applications
Cited by (107)
Signal based condition monitoring techniques for fault detection and diagnosis of induction motors: A state-of-the-art review
2020, Mechanical Systems and Signal ProcessingSmart classifier based prognostics and health management of induction motor
2020, Materials Today: ProceedingsCurrent signature analysis to monitor DFIG wind turbine generators: A case study
2018, Renewable EnergyCost-Effective IoT Extensions for Existing Public Coin Operated Washing Machine Towards Smarter Apartment Complexes
2023, Current Applied Science and TechnologyNovel Nonlinear High Order Technologies for Damage Diagnosis of Complex Assets
2022, Electronics (Switzerland)