The document describes a proposed advanced flood detection system based on sensor technology and machine learning algorithms. The system would use sensors to collect data on water levels and other environmental factors. This data would be processed using an Arduino and classified using a random forest machine learning algorithm to detect flooding. If flooding is detected based on threshold values, alerts would be sent to users through a cloud-based IoT platform to warn of the flood risk in real-time. The system is intended to provide low-cost flood monitoring and warnings.
Report
Share
Report
Share
1 of 19
More Related Content
FLOOD PPT 1.pptx
1. Advanced Flood Detection System based on Sensor
Technology and Machine Learning Algorithm
Guided By:
Prof. Kanchan Mahajan
Sandip Foundation's
Sandip Institute of Technology and Research Centre, Nashik
Department of Computer Engineering
Presented By:
• Akshay Jadhav [B150614244]
• Kshitija Khadke [B150614256]
• Shraddha Gite [B150614238]
2. OUTLINE
• INTRODUCTION
• OBJECTIVES
• PROBLEM STATEMENT
• SCOPE OF PROJECT
• ARCHITECTURE
• REQUIREMENTS
• ALGORITHM
• UML DIAGRAMS
• LITERATURE REVIEW
• REFERENCES
3. INTRODUCTION
• Floods are a major catastrophic event that allows for the catastrophic destruction of any
nation, affecting human lives and making outrageous harm to people and properties.
• Without proper monitoring and effective mitigation measures, these natural perils often culminate in
disasters that have severe implications in terms of economic loss, social disruptions, and damage to
the urban environment.
• The aim is to build an efficient flood warning system while maintaining reasonable production cost has
been a meaningful mission for many researchers
• In this system, a cluster of servers will collect and process data from the hydrological observation
stations in real time and with the help of Random Forest Algorithm for the classification of data the
available results can be displayed on client computer by remote access and issue warnings if there is a
risk of flood with the help of IoT.
4. OBJECTIVE
The main objective of this project is to develop and design a flood detection system that will detect
flood automatically and transmit data through IoT.
In the first stage, system will detect the current water levels on flood by taking sensor values from
outside environment and it will give real-time information to the appropriate station about severity.
At the second stage, the machine learning algorithm Random Forest is executed for classification
and examine the level of flood, information to evaluate if the level of water is typical or in unsafe
condition with the help of threshold values.
The proposed system is a low cost in design and easy for maintenance. This project will update the
water level at the web server and the system will issue an alert signal to the citizens for evacuation so
that fast and necessary action can be taken.
5. PROBLEM STATEMENT
• Many flood warnings stations have been developed
and installed in prosperous countries but the
manufacturing cost is usually too high to be practical
in developing countries.
• Therefore, building an efficient flood warning system
while maintaining reasonable production cost has
been a meaningful mission for many researchers
including our project.
6. SCOPE OF THE PROJECT
• Historical records have shown that flood is the most frequent natural hazard, accounting for 41% of all-natural
perils that occurred globally in the last decade. In this period alone (2009 to 2019), there were over 1566 flood
occurrences affecting 0.754 billion people around the world with 51,002 deaths recorded and damage estimated at
$371.8 billion. Put in context, these statistics only account for “reported” cases of large-scale floods.
• The ultimate goal was to improve the prediction accuracy, for this purpose some researchers have explored the
correlation among weather features and prediction accuracy and tried to find the best combinations of those features
to tune the performance.
• Few researchers on the other hand worked to train the mining technique well to achieve the high accuracy in
prediction. Few have compared the modern techniques with the conventional ones.
• Nevertheless, the current situation calls for improved ways of monitoring and responding to floods. The importance
of improved flood monitoring cannot be overemphasized given the growing uncertainty associated with climate
change and the increasing numbers of people living in flood-prone areas.
8. • To detect a flood the system observes various natural factors, which includes humidity, temperature, water level
and flow level. To collect data of mentioned natural factors the system consists of different sensors which collects
data for individual parameters.
• For detecting changes in humidity and temperature the system has a DHT11 Digital Temperature Humidity Sensor
and for the water level measurement it has water level sensor WL400.
• The system has a wi-fi connectivity, using the ESP8266 Wifi module, which connects the system to cloud; thus, it’s
collected data can be accessed from anywhere quite easily using IoT.
• All these sensors are then connected to ARDUINO UNO, which processes and saves data.
• The data movement from sensors to application can be viewed as a series of layered architecture that consists of
perception layer, network layer and application layer.
• The perception layer consists of various devices involved in getting the geological data. All of these data will be
transmitted to the application via wireless sensor network (WSN) and other communications equipment, the
network layer is responsible for handling this task. The data then received by the application layer is stored and
passed to the applications that need the data in order to do their tasks which is the application layer.
• The above-mentioned sensors measure the various environmental and weather-related parameters and monitor
them constantly. The data from these sensors is constantly fed to an Arduino controller. The Arduino program
checks for any irregularities in the sensor measurements and performs the associated computations. The Arduino
also has a Wi-Fi module attached to it, which enables it to send the sensor data to the remote IoT platform using
the IoT protocols over the Wi-Fi connection.
9. • The LCD is used to display the real-time values of the sensors. These data can also be viewed
on the cloud, which constantly retrieves the information from the remote IoT platform.
• If the value of any sensor crosses over a certain threshold value, an alert is sent to the end user
via the wifi module. Using this system, the flood-related parameters can be monitored from
anywhere in the world remotely.
• In this system we make use of an Arduino with sensors to predict flood and alert
respective authorities and sound instant alarm in nearby villages to instantly transmit
information about possible floods using IoT.
• All these features provided by the application can be efficiently used by any individual to
monitor the system. It is user friendly and avoids complication of different data used as the
user is only provided with what really is important.
10. REQUIREMENTS
Hardware
Arduino: The Arduino is the heart of the system all the sensors are connected to the Arduino and they
operate in a synchronized manner.
Wifi module: The ESP8266 is a System on a Chip (SoC), you get Wi-Fi communication, so you can use it to
connect to your Wi-Fi network, or connect to cloud.
Temperature and humidity sensor: DHT11 sensor for measuring temperature and humidity.
LCD display: 2x16 for displaying the data.
Ultrasonic sensor: HC-SR04 is used to measure the distance from the sensor to the water level.
Connecting wires, Bread Board
12. ALGORITHM
• Random forest, the "forest" it builds consists of a large
number of individual decision trees that operate as
an ensemble. Each individual tree in the random forest
spits out a class prediction and the class with the most
votes becomes our model’s prediction. (It can be used
for both classification and regression tasks)
• Random forest builds multiple decision trees and
merges them together to get a more accurate and
stable prediction.
• The reason that the random forest model works so
well is: A large number of relatively uncorrelated
models (trees) operating as a committee will
outperform any of the individual constituent models.
• The larger the number of trees, the more accurate the
result.
RANDOM FOREST