IRJET- Fuzzy Logic based Fault Detection in Induction Machines using Cloud

INTERNATIONAL RESEARCH JOURNAL OF ENGINEERING AND TECHNOLOGY (IRJET) E-ISSN: 2395-0056
VOLUME: 07 ISSUE: 02 | FEB 2020 WWW.IRJET.NET P-ISSN: 2395-0072
© 2020, IRJET | Impact Factor value: 7.34 | ISO 9001:2008 Certified Journal | Page 440
 Fuzzy Logic based Fault Detection in Induction Machines using Cloud
Madhuri P. Tale1
[1]Department of Electronics & Telecommunication Engineering, BAMU University, DIEMS,
Aurangabad, (M.S)India.
--------------------------------------------------------------------------------------***------------------------------------------------------------------------------------
Abstract - This paper presents monitoring system for an
induction motor based on Internet of Things (IOT) for
safe and economic data communication in industrial
fields. The main purpose of this paper are to monitor
fault analysis on an induction motor using experiments
as well as simulation along with failure identification
techniques applied for condition monitoring of the
motors and to design an on-line condition monitoring
system with fuzzy logic controller using cloud. In this
paper, work is divided in to two phases. The phase one
was modelling of single phase induction motor, in single
phase reference frame using Matlab/Simulink and
designing an intelligent system for condition monitoring
of the motors[4]. The phase two was implementation of
on-line condition monitoring system through cloud[1][6].
Keywords: - Cloud, Condition monitoring, Fuzzy logic,
Induction motor, Internet of Things (IoT),
MATLAB/Simulink, Raspberry Pi.
I. INTRODUCTION
Induction motor is the single most common
electromechanical energy conversion device available for
various industrials applications because of the reason is
the wide variety of characteristics like robustness, self
starting, high efficiency, low cost reliability, speed
control, flexibility etc[1]. An induction motor has two
electric circuit which are placed on the two main parts of
the machine: (i) the stationary part called the stator and
(ii) the rotating part called the rotor[2]. Power is
transferred from the part to the other by electromagnetic
induction. In this paper condition monitoring system for
induction motor has been developed in both simulation
model (Matlab/Simulink) as well as in real time (Cloud)
[3].The possible detection methods to identify the motor
faults are listed as follows.
1. Vibration measurement.
2. Noise measurement.
3. Temperature measurement.
4. Voltage measurement.
5. Current measurement.
The performance of the AC Induction motor depends
on on top of mentioned electrical, mechanical and
environmental parameters of the motor, in order
that the dominant strategies for prime performance
measure terribly sensitive to motor parameters. All
electrical, mechanical, environmental parameters like
current, voltage, speed, vibration, temperature, noise,
and external moisture of the induction motor are very
important for a drive system. The performance of an
induction motor is directly affected by the above
mentioned parameters. If any parameter of induction
motor crosses its cut off levels then quality of product
also changes, hence controlling the machines during the
process of production becomes a dangerous operation in
some specific industrial application. As an emerging
technology in modern wireless telecommunication
Internet of Things (IoT) has a lot of attention and to
provide many applications. The concept of “Internet of
Things" (IoT) is providing a best way for industrial
automation. In IoT each device are constituting a system
will be able to communicate with the other devices.
Which will help in industries to have better productivity,
management, safe environment and increased
throughout. Here in the proposed work the IoT is used
for monitoring and controlling the AC induction motor to
avoid the system failures.
II. DEVELOPED FUZZY LOGIC SCHEME
Fuzzy logic is a multivalued logic, allows to intermediate
values which is to be defined between conventional
evaluations like as yes/ no, true/ false etc. Fuzzy
controllers ar the foremost necessary application of fuzzy
theory. The result of induction motor condition made
based on fuzzy inference which is capable of giving
accuracy detection model. The structure of fuzzy abstract
thought system is shown in figure 1.
Fig 1. Structure of fuzzy Inference System

Fuzzyfication is defined as the conversion of crisp data in
to fuzzy data. After the fuzzy data goes to the block of
inference module. It is a rule base block .Rule or Logic
can be apply for the condition monitoring of motor to
obtain the logical output. After output of inference
module goes to the defuzzification block. This
Defuzzification block converts fuzzy output in to crisp
output.
Fig 2. Complete structure of Fuzzy Controller
From the above figure 2. Shows complete structure of
fuzzy controller. Fuzzy Inference System for motor fault
detection has created using Fuzzy Tool Box of MATLAB.
Fuzzy rules and membership functions are construct and
observing the data set. Speed is used as an input and
voltage is used for output. So for the measurements
related to the motor speed, more insight in to the data
are needed, so membership functions will be generated
for To Slow, Just Right, and Too Fast. For the
measurement related to the voltage condition will be
generated Less Voltage (Down), No Change, More
Voltage (Up). Membership functions are created by
observing the data set and the behaviour of stator
currents which are likely to cause faults in the motor. In
this study trapezoidal membership functions are used.
The membership functions for input (speed) and output
(voltage) variables are shown in figure3 and figure 4.
Fig3. Membership Functions of input motor Speed
Fig 4. Membership Functions for Output Motor voltage
Fig 5. Motor condition for membership function
Once the form of initial membership functions has been
determined, the fuzzy if-then rules can be add as shown
in above figure 5.
1. If motor speed is running too slow then more
voltage.
2. If motor speed is about right then no change voltage.
3. If motor speed is fast then less voltage. In this
paper, we've designated ranges for input and output
membership functions for predicting motor
condition whereas it's operational. For input
membership functions that is in this case for each input
stator speed we have selected range between 0 to 100.
Similarly for output membership functions that is voltage
of motor standing during this case is between 0 to 5.
There are various methods of defuzzification are
available. But during this paper we've
got utilized the centre of mass methodology for
defuzzification. The output of the fuzzy controller is used
as the command signal for the closed loop operation. If
any slight voltage unbalance occurs, then the output of

fuzzy inference system sets the output corresponding to
fault. Immediately the fault and the speed are stored in a
file for analysis purpose.
III. PROPOSED SYSTEM
The proposed system consists of Temperature sensor,
Vibration sensor, Speed (IR) sensor, Temperature sensor,
Current sensor, Voltage sensor for measurement circuits
and AC induction motor. The block diagram of proposed
system is below.
1 Phase
Induction
Motor
Vibration
Sensor
Noise
Sensor
Temp
Sensor
Voltage
Sensor
MCB
3008
(SPI)
Cloud
Raspberry
pi-3
Current
Sensor
Fig 6. Block diagram of proposed system
IV. SYSTEM WORKING
For induction motor parameter monitoring we are
using light weighted and easily configurable sensors like
Vibration sensor, Noise sensor, Temperature sensor,
Current sensor, Speed sensor, and Voltage sensor. All this
sensors are mounted on single phase motor. The single
phase induction motor has 230 V, 50 Hz power supply.
The single phase induction motor has to covert the
analog signal in to digital form. This is the main part of
hardware. The induction motor blocks goes to Vibration
sensor. It measure the vibrations of motor and they have
a transducer that converts mechanical force caused by
vibration. Noise sensor is the sensor which converts air
pressure vibrations due to sound in to electrical current.
The function of Temperature sensor is to measure the
temperature through an electrical signal. Voltage sensor
converts voltage measured between two point of an
electrical circuit in to a physical signal which is
proportional to the voltage. Then current sensor detects
and converts current to an easily measured output
voltage. Vibration sensor measure the heat of an object as
well as detects the motion. All sensor has analog output
hence it is connected to the input of the MCB 3008(Serial
Peripheral Interface). It is also serial peripheral interface.
The main function of MCB3008 IC is the analog to digital
converters. It has low cost 8 channel 10 bit A/D
convertor. All type of analog data converts in to digital
forms. Its output goes to the Raspberry pi- 3 board has
been used to for this research which has the ability to
acquire sensor data, and communicate with other
devices, store information in local, cloud server and alert
when fault is detected and display this message
according to values of sensors and also display values of
sensor i. e overload/over current, over voltage, noise
exceed, vibration exceed, temperature exceed.
Data that's obtained from the sensors area
unit transferred wirelessly to the native and cloud server
for analysis. The program has been set to process
real-time data and store it to the cloud with Thing speak
cloud computing platform. This saved data is available
from anywhere via internet. Figure 6. Shows block
diagram of the hardware connections.

Flow chart:-
Initializations
Read Sensor Information
Connect the raspberry pi to
wi-fi
Upload the information to
cloud (server)
Display the message with
reading & detect the faults
Read the commands from
cloud (Server
Check data on real time
simulation on Think speak
END
Fig 7: Flow chart of proposed system
The Proposed Algorithm are as follows:-
Step (0): Initialization: Initialize the username and
password which is supported to the python.
Step (1): Read the sensor information.
Step (2): Connect the Raspberry pi to wi-fi.
Step (3): Display the message with reading and detects
the fault in sensors.
Step (4): Upload information to cloud (Server).
Step (5): Read the command from cloud (Server).
Step (6): Check data on real time simulation.
Step (7): Check data on real time simulation on Think
speak.
Step (8): End.
V. RESULTS ANALYSIS
In this work one section industrial AC motor was used for
experimental purpose. The Sensor (vibration,
temperature, noise, current, voltage sensors) are
attached to the motor at right positions. Sensor data was
collected and processed using Raspberry pi-3 and
compared with the threshold values to avoid failure. The
experimental setup of the proposed system is as shown
in fig 8.

Fig 8: a) Setup of the proposed system
b) Motor display message with fault and reading
c) The above plots shows the real time simulation of the
variation of vibration, noise, temperature, current, and
voltage sensor with respect to date.
VI. CONCLUSION
This paper presents the idea of web of things for early
detection and watching of motor system failures. The
system has been designed to combine various parameter
measurement in real- time, improving the delectability of
different parameters namely vibrations, temperature,
speed, moisture, voltage and current consumption. The
concept of IoT is presented here controlling the motor.
The data received by the coordinator node is stored and
graphically presented in real- time by means of a
application in visual basics. The proposed system can be

easily upgraded to add other sensors on the sensing node
for the measurement of other parameters if required. The
system has a high autonomy, easy installation and low
maintenance costs. This is highly versatile technology for
condition monitoring and fault analysis of motors.
Acknowledgement
The authors wish to thank Dr. Ulhas Shiurkar, Director of
the DIEMS, Aurangabad for technical support. The
Authors are thankful to Dr. Rajesh Autee, HOD
Department of Electronics and Telecommunication
Engineering DIEMS, Aurangabad for their guidance.
REFERENCES
[1] W. T. Thomson " The Review of the On-Line Condition
Monitoring Techniques for the Three-Phase Induction
Motors Present and Future" Keynote address at IEEE
Symposium on Diagnostics for Electrical Machines and
Power Electronics and Drives, Gijon, Spain, pp on 3-18
Sept. 1999.
[2] Y.E. Zhongming and WU Bin, Ryerson "A Review on
Induction Motor Online Fault Diagnosis" Proceedings of
third international conference on Power Electronics and
Motion Control, 2000. IPEMC 2000.
[3] M.E.H. Benbouzid, H. Nejjari, "A Simple Fuzzy Logic
Approach for Induction Motors Stator Condition
Monitoring" IEEE Conference, 2001. IEMDC 2001.
Volume, Issue, 2001.
[4] Ramazan Bayindir, Ibrahim Sefa, Ilhami, Askin
Betkas, "Fault Detection And Protection of Induction
Motors Using Sensors" IEEE Transactions on Energy
Conversion, VOL.23, NO. 3, on September 2008. Pp.
734-741.
[5] R.Saravan kumar, K. Vinoth Kumar, Dr. K. K. Ray
“Fuzzy Logic Based Fault Detection in Induction
Machines Using LabView", IJCSNS International Journal
of Computer Science and Network Security, Vol.9, NO. 9,
on September 2009.
[6] A.K.Cakir "Remote Controlling And Monitoring of
Induction Motors Using Internet", faculty of technology,
department of electrical and electronic engineering, is
parta turkey(2014).

IRJET- Fuzzy Logic based Fault Detection in Induction Machines using Cloud

More Related Content

What's hot (18)

Similar to IRJET- Fuzzy Logic based Fault Detection in Induction Machines using Cloud (20)

More from IRJET Journal (20)

Recently uploaded (20)

IRJET- Fuzzy Logic based Fault Detection in Induction Machines using Cloud