1.

1 Internet of Things (IoT)

An IoT environment comprises various sensors, communication media, and devices that interact and
exchange data with each other. These devices utilize the cloud to exchange sensor data and process it
for decision-making [1]. The devices operate autonomously, as depicted in Figure 1 [2], and users can
interact with them.

To illustrate an IoT environment, diverse IoT-based technologies transmit data to a central hub for data
collection. This data can be processed for analysis or sent to different interfaces, such as smartphones or
man-machine systems, to perform desired actions. This represents a significant technological
advancement that enables more accurate and efficient interactions between humans and machines.
Research indicates that by 2025, there will be a billion interconnected devices on the internet. The
connectivity of devices creates intelligent networks, and smart devices are being developed. Once
connected, data can be analyzed in various ways.

By combining IoT and automation, better and faster products can be created at lower costs. Through
automation, operations can be performed automatically and with greater precision. IoT provides
opportunities to optimize resource consumption while saving time and money.

1.1.1 Benefits of IoT

The emergence of IoT and automation has accelerated the monitoring of comprehensive systems and
processes. Moreover, by adding value and reliability, the Internet of Things enhances the consumer
experience when using products and services. IoT also enables businesses to save time and money by
implementing surveillance systems. These monitoring systems can be employed in distribution networks
to reduce downtime after a failure. Deploying IoT-based devices in offices and homes will expedite
various aspects of daily life, as self-sufficient systems like thermostats, air conditioners, and lighting
allow people to allocate more time to other responsibilities.

IoT-based systems can be remotely controlled and easily monitored, even in adverse weather conditions
and challenging work environments. Consequently, IoT-based technologies offer enhanced reliability,
speed, accuracy, cost-effectiveness, and security in their operations.

1.2 Applications of Internet of Things (IoT) 1.2.1 Application of IoT in Energy Systems

In the context of the Australian energy market regulated by AEMO (Australian Energy Market Operator),
the implementation of IoT has been successful. The Internet of Things provides real-time data and
allows market participants to monitor various data points.

The IoT system enables different distribution companies to synchronize and operate under a single
management entity. By deploying sensors throughout the distribution network, troubleshooting can be
minimized or even prevented. Additionally, smart energy meters facilitate faster billing processes. These
sensors can also communicate with each other to regulate the flow of electricity based on the maximum
load permitted for a particular user. In contrast, IoT sensors can adjust gas pressure in accordance with
regulations. Since both the gas and electricity distribution systems are centrally controlled,
communication methods can be standardized for improved efficiency and coordination.

1.2.2 IoT-Based Monitoring of Transmission Systems

Electricity lines play a crucial role in IoT-based surveillance systems. However, they are susceptible to
various natural hazards that can impact the stability and reliability of the system [3]. While there are
currently several wireless network systems in practical use, they often face challenges associated with
high operational and maintenance expenses.

1.3 Transformer Protection System

This project focuses on the continuous monitoring of oil temperature in a transformer to prevent
overheating and potential faults. Transformers are crucial components of the power transmission and
distribution infrastructure, and monitoring them is essential to avoid costly repairs or replacements, as
well as service disruptions and revenue loss.

To monitor the oil temperature, a system has been developed that senses the temperature and
automatically activates the circuit breaker if the temperature exceeds the specified limit. It also alerts
the electricity department through a buzzer alarm system and displays the temperature on an LCD
display.

The temperature is sensed using an inexpensive method where the Arduino UNO reads the analog data
from an ADC module and converts it into the actual temperature reading. The Arduino UNO is
programmed to determine the appropriate action based on the temperature input and displays the oil
temperature on the LCD display.

The objectives of this project include continuous oil temperature reading, alerts through a buzzer, visual
alerts and continuous monitoring on the LCD display, and reliability for industrial needs.

The project provides exposure to various aspects such as initializing the ADC module of Arduino UNO,
understanding temperature sensor characteristics, embedded C programming, PCB designing, and LCD
interfacing with Arduino UNO.

The major building blocks of the project include a regulated power supply, Arduino UNO, temperature
sensor, relay with driver, buzzer, LCD display with driver, and LED indicators.

The software used for hardware simulation is Proteus, while the Arduino IDE software is used for
programming the Arduino UNO.

Arduino:
Arduino is a microcontroller board designed to facilitate the development of interactive objects and
environments. It features open-source hardware and is built around either an 8-bit Atmel AVR
microcontroller or a 32-bit Atmel ARM. The board includes various hardware components such as a USB
interface, 6 analog input pins, and 14 digital I/O pins, allowing users to connect different extension
boards.

The Arduino Uno board, specifically, is based on the ATmega328 microcontroller. It offers 14 digital I/O
pins, with 6 of them capable of functioning as PWM outputs. The board also includes a 16 MHz ceramic
resonator, an ICSP header, a USB connection, 6 analog inputs, a power jack, and a reset button. These
features provide all the necessary support for the microcontroller. To begin using the Arduino Uno, it
can be connected to a computer via a USB cable or powered using an AC-to-DC adapter or battery.
Unlike other boards, the Arduino Uno does not utilize the FTDI USB-to-serial driver chip. Instead, it
features the Atmega16U2 (or Atmega8U2 in older versions) programmed as a USB-to-serial converter.

There are various types of Arduino boards available, including third-party compatible versions. The
official versions are the Arduino Uno R3 and the Arduino Nano V3. Both of these boards are equipped
with a 16 MHz Atmel ATmega328P 8-bit microcontroller, 32KB of flash RAM, 14 digital I/O pins, and six
analog I/O pins. It's important to note that the 32KB of flash RAM is not intended for running operating
systems like Windows. Arduino projects can be standalone or can communicate with software running
on a computer, such as Flash, Processing, or Max/MSP. The board is powered by a 16 MHz ceramic
resonator and includes a USB connection for power and communication. Additionally, it is possible to
expand the storage capacity by adding a micro SD/SD card.

Features of the Arduino Uno Board:

1. It offers an easy-to-use USB interface, allowing for seamless connection to a computer as a
virtual serial port. This enables simple and reliable serial communication between the board and
the computer.

2. The board utilizes the ATmega328 chip as its microcontroller brain, which provides a wide range
of hardware features including timers, external and internal interrupts, PWM pins, and multiple
sleep modes. These features enhance the capabilities of the board for various applications.

3. Being an open-source design, Arduino benefits from a large community of users who contribute
to its development and troubleshooting. This active community support makes it easier to find
assistance and resolve issues when working on projects.

4. The board operates at a clock speed of 16 MHz, which is sufficient for most applications without
unnecessarily speeding up the microcontroller.

5. Arduino boards come with built-in voltage regulation, making power management convenient.
They can be powered directly from a USB port without the need for an external power source.
Additionally, an external power supply of up to 12V can be connected, and the board regulates
it to provide both 5V and 3.3V power.
6. The board features an In-Circuit Serial Programming (ICSP) connector, which allows for direct
interfacing with the Arduino as a serial device, bypassing the USB port. This connector is useful
for re-bootloading the chip in case of corruption or when the USB port is not accessible.

7. Arduino boards have 32 KB of flash memory available for storing code, providing ample space
for programming and storing instructions.

8. An on-board LED is connected to digital pin 13, which serves as a convenient debugging tool,
allowing for quick visual feedback during code development. Additionally, a reset button is
provided to easily restart the program on the chip.

9. Arduino was created in 2005 by two Italian engineers, David Cuartillas and Massimo Banzi, with
the goal of teaching students how to program microcontrollers and enhance their electronics
skills for real-world applications.

10. Arduino Uno, like other Arduino boards, can sense its environment through various sensors and
control external devices such as lights and motors. The programming language used for Arduino
is based on Wiring, and the development environment is based on Processing, making it
accessible and user-friendly for beginners.

11. LM35 Temperature sensor

Temperature is a commonly measured environmental parameter due to its
impact on various systems such as physical, chemical, electronic, mechanical,
and biological systems. Different systems and processes have optimal
temperature ranges for their efficient operation. Temperature sensing can be
achieved through direct contact with the heat source or remotely using non-
contact methods that rely on measuring radiated energy.

In the market, there is a wide range of temperature sensors available to suit

different applications. Some commonly used types of temperature sensors
include:

Thermocouples: These sensors utilize the principle of thermoelectric effect to

measure temperature. They consist of two different metal wires joined at a
junction. The temperature difference between the junction and the other ends
of the wires generates a voltage that is proportional to the temperature.

Resistance Temperature Detectors (RTDs): RTDs are temperature sensors that

rely on the change in electrical resistance of certain materials with temperature.
Typically, platinum is used as the sensing element due to its stable and
predictable resistance-temperature characteristics.

Thermistors: Thermistors are semiconductor devices whose electrical resistance

varies with temperature. They are typically made from ceramic or polymer
materials and exhibit a highly nonlinear response to temperature changes.

Infrared (IR) Sensors: Infrared temperature sensors measure temperature

without direct contact by detecting the thermal radiation emitted by objects.
They can be used for both non-contact and remote temperature measurement.

Semiconductor Sensors: Semiconductor temperature sensors, such as integrated

circuit (IC) temperature sensors, utilize the temperature-dependent
characteristics of semiconductor materials to measure temperature. These
sensors are often integrated into electronic devices and provide accurate and
reliable temperature readings.

The choice of temperature sensor depends on factors such as the required

temperature range, accuracy, response time, and the specific application. By
 selecting the appropriate temperature sensor, precise temperature monitoring
can be achieved in various systems and processes.

Working
A temperature control circuit can be created using an Arduino Uno and an LM35 temperature sensor,
along with other necessary components. A 16×2 LCD is utilized to display the current temperature and
set points. The LM35 sensor provides an analog output that is proportional to the temperature, which is
connected to the Arduino's analog input A0. The temperature reading is then compared to the set
points. If the temperature exceeds the set point, indicating that it is too high, the heating element (e.g.,
a relay-connected heater) is turned off. Conversely, if the temperature is below the set point, the relay
(heater) is turned on. The status of the heater (on/off) is displayed on both an LED and the LCD. Two
tactile switches are used to adjust the temperature set point.

In this setup, the Arduino functions as the main controller for the digital temperature control system. It
receives input from the temperature sensor, compares it to the predefined set points, and controls the
heating device accordingly. The system also provides a display of the device's status (on/off) and the
current temperature. The temperature controller ensures precise temperature regulation by comparing
the actual temperature, measured by the LM35 temperature sensor, to the desired control temperature
(set point). Based on this comparison, the controller sends an output signal to control elements such as
heaters or fans.

Using the LM35 as a temperature sensor, the circuit can measure the temperature of a transformer or
any other target object. Since the LM35 provides an analog value, it needs to be converted to a digital
value for comparison with a given temperature range. The Arduino is responsible for converting the
analog value to a digital format. The recorded temperature value is then displayed on a digital display.

In this setup, if the recorded temperature exceeds the given temperature range, an alarm or buzzer is
activated to indicate that the transformer's temperature has exceeded its rated capacity. This
information can be used to take necessary actions. Additionally, the circuit breaker can be operated to
interrupt the loads connected to the transformer, helping to maintain the temperature within the rated
capacity and ensuring the transformer's health and longevity.

Conclusion:

We have utilized a combination of an LCD and the LM35 temperature sensor to create a straightforward
temperature controller using Arduino. When the temperature applied to the LM35 sensor exceeds the
set point, the relay (Heater) is activated. Conversely, if the temperature is below the set point, the relay
is turned off. By employing the LM35 temperature sensor and Arduino, we can monitor the temperature
of the transformer and establish a specific temperature range within which the transformer should
operate to stay within its rated capacity.
This protective circuit provides an indication if the temperature of the transformer surpasses the
predetermined temperature range, enabling us to take necessary measures to prevent overheating. For
instance, we can utilize a circuit breaker and relay circuit to interrupt the loads connected to the
transformer.

Due to the minimal and affordable components required for this protective circuit, such as the LM35,
Arduino, Buzzer/LED, Relay, and Circuit Breaker system, the overall cost of the circuit is significantly
reduced. The protective circuit is easy to comprehend, operate, and maintain, further minimizing project
expenses. Additionally, this circuit can help reduce faults and losses associated with the transformer,
promoting efficient operation. Thus, this system provides an accurate temperature measurement, and
the protective circuit responds accordingly. Consequently, this protective circuit safeguards the
transformer from overheating and enhances its health and efficiency.

2.1 Problem Formulation

Distribution agencies in the power supply market face the challenge of providing consistent and reliable
electricity to both residential and commercial customers. Service transformers are installed in remote
and inaccessible locations, making preventive and routine maintenance difficult for the distribution
utility staff. With a large number of these transformers in operation, maintenance is often conducted
reactively after a failure occurs. To delay the need for capital investment in maintaining modern
facilities, distribution agencies are opting to utilize existing equipment that is at the higher end of its
capacity range. This approach allows them to prioritize investments and postpone the expenses
associated with maintaining and upgrading infrastructure.

2.1 Literature Survey

The main objective of IoT is to connect real-world devices and create an intelligent digital environment.
The literature has discussed the functioning of IoT and its practical applications, including technologies
such as NFC, Zigbee, WiFi, Bluetooth, etc., which are commonly employed in IoT implementations. These
technologies enable the establishment of a large virtual network through peer-to-peer connections
among diverse devices.

Transformer condition assessment methods have been extensively explored in the literature. There is a
growing need for non-invasive diagnostic and manipulation tools to evaluate the internal state of
transformers. The causes of transformer failures have been identified by the CIGRE operating
organization. Research has focused on power transformer reliability issues, covering areas such as
comprehensive storage, gas-in-oil analysis, interpretation techniques, and failures analysis.

Various techniques related to sample storage, gas-in-oil extraction, and analysis have been studied, as
they are crucial for the diagnostic significance of gas-in-oil analysis results. Utility experience with power
transformer and reactor faults, as well as non-fault scenarios, has also been investigated. Dissolved Gas
Analysis (DGA) is recognized as a potent diagnostic tool, capable of identifying issues related to thermal,
chemical, and dielectric aging. Other methods, including the examination of operational conditions, flow
response analysis, partial discharge analysis, and UHF partial discharge measurement, have been
explored for transformer monitoring and diagnostics.

Research has been conducted on the development and application of transformer fault detection
systems, such as the Transformer Inter Turn Fault Detection System (TIFDS), which eliminates the need
for secondary-side current transformers. Additionally, a method utilizing harmonic analysis of
transformer no-load and light load currents has been proposed for the early detection of inter-turn
faults.

Current and voltage sensors have been used in algorithm development for identifying inaccuracies and
winding conditions, even in the presence of OLTC online tap changer issues. Symmetrical component-
based methods have been discussed for sensitive and reliable fault detection between windings and
extreme short circuits. Fuzzy set theory has been employed to calculate the Health Index (HI) for oil-
immersed transformers, requiring offline techniques such as DGA and furan analysis. The determination
of HI for transformers and the integration of various transformer tests' results have also been proposed.

The research emphasizes the implementation and utilization of HI determination in transformers, where
anomaly-related data is sent via SMS to the asset owner or application engineer. The effectiveness of
this approach lies in monitoring and predicting the health state of distribution transformers by
forecasting health index levels using the information provided in Table 3.

3.1.1 Cloud Server

The machine employed in this project utilizes the ThingSpeak platform for data processing and analytics.
ThingSpeak enables the retrieval of data from devices in the form of short messages through the HTTP
protocol and the internet. It provides real-time visualization of uploaded data, location tracking, and
integration with social media. Moreover, ThingSpeak offers a MATLAB tool that allows data processing
either on an all-in-one data server or in the cloud. In the proposed system, 3000 bytes of data are sent
to ThingSpeak every 15 seconds.

3.1 Workflow of the proposed system

The proposed system follows the following workflow:

a) Acquire sensor-based data

The system collects data about transformers using various sensors. To implement this, the system
utilizes the free ThingSpeak communication libraries for embedded devices, implemented in C-sharp.

b) Upload data to the cloud and visualize it

The collected sensor-based data is uploaded to the cloud and displayed accordingly. The data is
uploaded to the ThingSpeak server, as shown in Figure 13.

c) Create an algorithm for the health index

An algorithm for calculating the health index is developed using MATLAB code. The MATLAB code
includes a timer function with a predefined rate and a period of 60 seconds.

Below is the code for calculating the health index based on the data retrieved from the ThingSpeak
server. In this example, the MATLAB programming language utilizes the thingSpeakRead function to
read data from ThingSpeak.

System architecture overall

The design of the entire machine is divided into components:

1) Hardware implementation and design
2) Programming software program

3.2 Hardware implementation and design

The hardware components used in the implementation are as follows:

a) Current Transformer

b) Rectifier-based voltage sensor

c) Temperature sensor

d) Analog to digital converter (ADC)

e) Ethernet shield

f) Internet connection

a) Current Transformer

To measure load current variations, a secondary current transformer with a CT ratio of 50A/10mA is
used. Transformer overloads are anticipated when the load exceeds 90% of its capacity. The secondary
current of the transformer is required to calculate the harmonic load current, which generates excessive
heat and winding losses during overloading. A parallel resistor connected across the secondary terminals
of the CT converts the current into voltage.

b) Rectifier-based voltage sensor

A voltage variation of ±6% is allowed for distribution transformers. Flux variation in the transformer
leads to a proportional voltage change, and overvoltage causes increased iron losses. The circuit diagram
for the rectifier-based voltage sensor is shown below. The circuit includes a filter circuit of 700uF, a
series resistor of 5K ohm 20W connected to a zener diode, and LEDs to discharge capacitors during
voltage fall.

c) Temperature sensor

A temperature sensor is used to monitor the winding temperature, which is a critical parameter in
determining the transformer's load capacity. The LM35 temperature sensor is employed due to its
temperature range of -55℃ to 150℃, and literature suggests that the winding temperature can exceed
60℃.

d) Analog to digital converter (ADC)

An ADC based on ATMEGA 2560, specifically the Arduino Mega, is utilized to convert analog data from
the sensors into a digital format. The Arduino Mega has a clock speed of 16MHz, 128KB flash memory,
and 8KB RAM.
e) Ethernet Shield and Internet Data Card

An Arduino-based Ethernet shield is used, connected to a Wi-Fi router with a 3G data card for internet
connectivity.

General Circuit Diagram

The following diagram represents a standard circuit diagram for a 3-phase distribution transformer
system.

Hardware Configuration and Testing

The ATMEGA 2560-based Arduino Mega is configured to collect readings from all the sensors and upload
them to a web server using an Ethernet-based internet connection. The microcontroller takes readings
and uploads them at regular intervals of 15 seconds. A resistive load is used as the load for the
transformer.

3.1 Software coding

Software coding is divided into two parts:

a) Arduino Mega

The coding for the microcontroller is done using the Arduino IDE. The IDE, or Integrated Development
Environment, is a software running on the computer that allows for programming the microcontroller.
Arduino programming is based on the C language.

b) MATLAB software
MATLAB software is used on the server side. It processes data every 15 minutes from the web server
based on the set of rules designed, which calculates the health index of the transformer.

All-in-one server algorithm

The algorithm flow chart for the all-in-one server is depicted in Figure 15.

Hardware Setup

The hardware setup for the IoT-based system is shown in Figure 21. It includes the complete circuitry, a
web modem with a 3G modem, and a microcontroller.

Testing and Result

Online parameters of a single-phase transformer are presented in Figure 22. Area chart 1 displays
temperature variations, area 2 shows voltage variations, and area 3 displays current variations.

Figure 23 represents all the measured parameters uploaded by the client into the cloud server, while
Figure 24 displays the calculated health index by the all-in-one server, which is 20%.

Health Index Categorization

The fitness index calculated above is classed based at

the limits proven in discern 6. As shown in figure 24, the fitness index
calculated above is 20% so in step with class, our machine is in state of
failure any time.

Discussion:

The presented IoT system is an advanced and improved model based on GSM technology. Unlike other
systems that rely on third-party clouds or general-purpose servers, this system eliminates those
limitations and introduces a new opportunity to update the algorithm periodically without incurring
additional costs. Additionally, it offers the following advantages:

1. Cost-Effectiveness: The system eliminates the need for expensive third-party cloud services,
making it a more cost-effective solution.

2. Algorithm Updates: The system allows for easy algorithm updates, providing flexibility and
adaptability to changing requirements without incurring extra expenses.

3. Independence: By removing the reliance on external servers, the system ensures independence
and self-sufficiency in data processing and analysis.

Application:

The introduced system has wide-ranging applications in monitoring distribution transformers under
substations. Given the large number of substations and the geographical dispersion of distribution
transformers, this system can be effectively utilized to autonomously monitor all the distribution
transformers within a specific area under a substation.

Distribution agencies typically manage numerous substations, which are often located in remote areas
far from the utilities' central headquarters. Additionally, these substations are spread across different
geographical locations. The system's ability to monitor and track distribution transformers in such
settings makes it highly applicable and valuable in ensuring the efficient and reliable operation of the
power distribution network.

By implementing this system, distribution agencies can achieve comprehensive monitoring and
management of their distribution transformers, enabling timely detection of faults or abnormalities and
facilitating proactive maintenance and troubleshooting. This leads to enhanced operational efficiency,
reduced downtime, and improved overall performance of the distribution network.

Future Scope:

1. The system has great potential for application in transmission lines by incorporating various
communication protocols such as LoRa, RF 434MHz, etc. This would enable efficient monitoring
and management of transmission infrastructure, improving the overall reliability and
performance of the power transmission system.

2. Integration of genetic algorithms can enhance the accuracy of the health index calculation by
identifying and compensating for faulty sensor readings. This advanced algorithmic approach
 would contribute to more reliable and precise monitoring of transformer health, even in the
presence of sensor faults.

3. The system can be expanded to cover large campuses or residential societies that have multiple
acute substations. By implementing the system across such extensive areas, the operation and
monitoring of substations can be conducted remotely, leading to improved efficiency and cost-
effectiveness.

4. The system can play a crucial role in reducing post-fault clearing time in the distribution
network. By promptly detecting faults and providing real-time alerts, the system enables faster
response and maintenance, minimizing downtime and ensuring a more resilient power
distribution infrastructure.

5. There is also potential to upgrade the system to accommodate three-phase transformers,

enabling more accurate and comprehensive monitoring of the health index for such
transformers. This expansion would further enhance the system's capabilities and applicability in
a wider range of power distribution scenarios.