On-Line Monitoring and Classification of Stator windings Faults in Induction Machine Using Fuzzy Logic and ANFIS Approach Abstract-- the induction machines drives becomes more and more important used in many industrial applications. Their attractiveness is largely due to their simplicity, ruggedness and low cost manufacture, easy maint00enance, high power efficiency and high reliability, are susceptible to various types of electrical and/or mechanical faults that can lead to unexpected motor failure and consequently impulsive downtime. This made necessary the monitoring function condition of these machines types for improved an exploitation of the industrial processes. The aim of this task is the proposal of a monitoring strategy based on the fuzzy logic inference system (FIS) and the neuro-fuzzy inference system (ANFIS) for monitoring and classification of electrical faults types, especially the open phase and inter- turns short-circuit in the stator windings. The principle adopted for the strategy suggested is based on monitoring of the average root mean square value of stator current (RMS). Mathematical models and simulations results are presented to validate the efficiency of this approach. Index Terms-- Monitoring; Classification; FIS; ANFIS; RMS. 1. INTRODUCTION Different of electrical machines types are present in several processes and industrial equipments. But the induction machines are currently the principal means in the industrial sector for conversion electrical energy into mechanical driving and they are play important roles in various industrials processing. Though their low cost, simple maintenance, from the reliability and robustness perspective point [1, 2]. Although all these advantages, these machines are easily prone to failure since are frequently installed in variety and the hostile environment that may be easily led to the deterioration. Moreover, several problems may occur during their function because of thermal, mechanical and electrical stresses, incorrect functioning condition or manufacturing defects [3]. In recent years the online monitoring and diagnosis techniques of faults found in three-phase induction machines are study under various approaches by many research tasks, since of its considerable interest for the continuity of the industrial processes service [4, 5].In specifically the most common electrical faults in induction machines are related to the stator windings, as inter-turn shorte circuit account for more than 30% of all faults, also the open stator phase default is one of faults in stator [7, 8] Early faults detection allows to minimize the downtime, the turn-around time of the process in question, to avoid the damaging consequences, and to reduce the financial losses [9]. The majority of the monitoring approaches are based on the analysis of electromagnetic magnitude such that the magnetic flux, the stator or rotor current, and the neutral voltage [10, 11]. In this case, by measuring accessible and easily quantifiable magnitudes, includes the stator currents of the induction machine for calculate their RMS values to analyze them in a minimum of time and to conclude the state of the induction machine [12]. However, through this work, we will be interested particularly in the open circuit and short-circuit inter-turns faults in stator winding of the induction machine (IM). The inter-turn short circuit fault in stator windings can propagate and can be developed either due to total defect insulation inter-turns of stator winding, leading to phase to ground or phase to phase faults. Some importance is therefore attached to the early detection of stator faults [13, 14]. So, the approach that we propose is based on the fuzzy logic inference system (Fis) and Adaptive Neuro-Fuzzy System Inference (ANFIS), in order to increase the efficiency and the reliability of the on-line monitoring and classification faults in the supervision of the induction machine [15, 16]. The models of the approach as well as the global model are simulated by using software MATLAB ® /SIMULINK and the obtained results of simulations in a healthy function and short-circuit or open phase faults are presented and interpreted. Merabet. Hichem Research Center in Industrial Technologies (CRTI) P.O. Box 64, Cheraga, Algeria. [email protected]Bahi. Tahar Electrical Department, University of Annaba, Algeria [email protected]Drici. Djalel Research Center in Industrial Technologies (CRTI) P.O. Box 64, Cheraga, Algeria. [email protected]Bedoud. Khouloud Research Center in Industrial Technologies (CRTI) P.O. Box 64, Cheraga, Algeria. [email protected]Boudiaf. Adel Research Center in Industrial Technologies (CRTI) P.O. Box 64, Cheraga, Algeria. [email protected]Proceedings of the International Conference on Recent Advances in Electrical Systems, Tunisia, 2016 ISBN: 978-9938-14-953-1 (187) Editors: Tarek Bouktir & Rafik Neji
6
Embed
On-Line Monitoring and Classification of Stator windings ...journal.esrgroups.org/jes/icraes/CDICRAESFinal/ICRAES16ProcPaper… · On-Line Monitoring and Classification of Stator
This document is posted to help you gain knowledge. Please leave a comment to let me know what you think about it! Share it to your friends and learn new things together.
Transcript
On-Line Monitoring and Classification of Stator windings Faults in
Induction Machine Using Fuzzy Logic and ANFIS Approach
Abstract-- the induction machines drives becomes more
and more important used in many industrial applications.
Their attractiveness is largely due to their simplicity,
ruggedness and low cost manufacture, easy maint00enance,
high power efficiency and high reliability, are susceptible to
various types of electrical and/or mechanical faults that can
lead to unexpected motor failure and consequently impulsive
downtime. This made necessary the monitoring function
condition of these machines types for improved an
exploitation of the industrial processes. The aim of this task
is the proposal of a monitoring strategy based on the fuzzy
logic inference system (FIS) and the neuro-fuzzy inference
system (ANFIS) for monitoring and classification of
electrical faults types, especially the open phase and inter-
turns short-circuit in the stator windings. The principle
adopted for the strategy suggested is based on monitoring of
the average root mean square value of stator current (RMS).
Mathematical models and simulations results are presented
to validate the efficiency of this approach.
Index Terms-- Monitoring; Classification; FIS;
ANFIS; RMS.
1. INTRODUCTION
Different of electrical machines types are present in
several processes and industrial equipments. But the
induction machines are currently the principal means in
the industrial sector for conversion electrical energy into
mechanical driving and they are play important roles in
various industrials processing. Though their low cost,
simple maintenance, from the reliability and robustness
perspective point [1, 2].
Although all these advantages, these machines are
easily prone to failure since are frequently installed in
variety and the hostile environment that may be easily led
to the deterioration. Moreover, several problems may
occur during their function because of thermal,
mechanical and electrical stresses, incorrect functioning
condition or manufacturing defects [3].
In recent years the online monitoring and diagnosis
techniques of faults found in three-phase induction
machines are study under various approaches by many
research tasks, since of its considerable interest for the
continuity of the industrial processes service [4, 5].In
specifically the most common electrical faults in induction
machines are related to the stator windings, as inter-turn
shorte circuit account for more than 30% of all faults, also
the open stator phase default is one of faults in stator [7, 8]
Early faults detection allows to minimize the
downtime, the turn-around time of the process in question,
to avoid the damaging consequences, and to reduce the
financial losses [9].
The majority of the monitoring approaches are based on the analysis of electromagnetic magnitude such that the
magnetic flux, the stator or rotor current, and the neutral
voltage [10, 11]. In this case, by measuring accessible and
easily quantifiable magnitudes, includes the stator currents
of the induction machine for calculate their RMS values to
analyze them in a minimum of time and to conclude the
state of the induction machine [12].
However, through this work, we will be interested
particularly in the open circuit and short-circuit inter-turns
faults in stator winding of the induction machine (IM).
The inter-turn short circuit fault in stator windings can
propagate and can be developed either due to total defect insulation inter-turns of stator winding, leading to phase to
ground or phase to phase faults. Some importance is
therefore attached to the early detection of stator faults
[13, 14].
So, the approach that we propose is based on the fuzzy
logic inference system (Fis) and Adaptive Neuro-Fuzzy
System Inference (ANFIS), in order to increase the
efficiency and the reliability of the on-line monitoring and
classification faults in the supervision of the induction
machine [15, 16]. The models of the approach as well as
the global model are simulated by using software
MATLAB®/SIMULINK and the obtained results of
simulations in a healthy function and short-circuit or open
phase faults are presented and interpreted.
Merabet. Hichem
Research Center in Industrial Technologies (CRTI) P.O.