Smart Dam Monitering & Controling

SOCIAL BENEFICIAL PROJECT: SMART DAM
[Team zombies]
Team Details
Participant
Name
CT /DT
Number
Role (Team
Leader / Member)
Bachelors Discipline
Expected Year
of Passing
Gender
KRISHNENDU
DATTA
CT20162
048187
Team Leader Electrical Engineering 2020 M
SUBHENDU
GHORAI
CT20162
050714
Team Member Mechanical Engineering 2020 M
SUTANU
MONDAL
CT20162
050739
Team Member Mechanical Engineering 2020
M
SANJIB
BETAL
CT20162
050735
Team Member
Computer Science and
Engineering
2020 M
Page 1

SOCIAL
1. Problem Statement
After analyzing various reports on dam incidents, dam failure, flood occurrence, dam
alerts and irrigation issues due to drought we identified a high-priority call for real time
dam water level monitoring and prior alerting and control system which ensures the
public safety and smart utilization of reservoir water.
2. Use Case Overview
The proposal is titled “SMART DAM WATER MONITORING AND
CONTROLLING” and the objective of the project is to make the existing system smart
by adding connectivity, Artificial intelligence (AI), IOT, cloud and dashboard.
3. Use Case Description
A dam is nothing but a barrier constructed to hold back water and raise its level, forming
a reservoir used to generate electricity or as water supply for various activities such as in
irrigation.
In present scenario most dams are manually monitored and data are sent via traditional
modes, manual observation and transmission results in a time loss between the data
observed in dam site and decision taking level. This sometimes causes loss of worthwhile
real time data.
During floods the dam is subjected to heavy volume of water, in such alarming situation
manual operation becomes very risky and any human error can put the lives of thousands
of people in danger.
One of the major problems is that opening and closing of floodgate is manually operated
and there is no intelligent estimation of required volume of water to be released.
Water being an important resource for living, it needs to be conserved and preserved.
Therefore its distribution and usage is of utmost consideration.
Page 2

The three major stakeholders involved in this system are dam authorities, researchers
and the common people. These are the concerned people considered in our solution.
3.1 Current system
Current system has no inclination towards the real time data available to dam authorities,
common people and the researchers. There is no proper monitoring and controlling
procedures. Let’s consider the issues of the three main stakeholders as discussed in the
previous section.
 DAM AUTHORITIES
1. Dam monitoring is done through traditional surveillance techniques
and becomes excessively risky during bad weather.
2. Doesn’t have real-time view of different parameters therefore there
is a time lag in providing the data observed in dam site to the
decision makers.
3. During both flood and draught situations, decision to open or close
the water gate is a censorious action that needs to be undertaken as
soon as possible, late decision will not only cause flood
downstream but will also damage the structure.
 Researchers
1. Researchers want observed data to be readily available for research
purpose as well as monitor the authentic time changes in various
parameters.
2. Dam parameters’ data collection is mostly unavailable in present
systems.
 Common people/Farmers
1. Unenlightened about the parameters like rainfall, Dam water level
and gate status.
2. Uncertainty about water for crops, sudden rise of backwater and
sometimes flood.
Page 3

3.2 Proposed solution
Proposal is to make the complete system smart. Considering the issues listed fixing them
the best way possible is considered. In short following is done for each of the
stakeholders.
 DAM AUTHORITIES
1. A dashboard is provided with all vital parameters sensed from
different sensors and analytics of these parameters along with
weather forecasted from various sources.
2. Real time monitoring with the aid of IOT and AI removes the time
lag between data observation and decision making.
3. Flood Gate opening and closing time is estimated with the help of
Intelligent Decision support system (IDSS) based on hydrological
parameters like water level, rain fall and gate position.
 Researchers
1. With the help of cloud and IOT the various dam parameters are
readily available for research purpose.
2. Researchers can monitor the authentic time changes in various
parameters.
 Common people/Farmers
1. Installation of water alarm systems in the downstream region of
dams to warn the authorities and to alert the population for ensuring
the evacuation of the flood prone area.
The water alarm system consists of sirens that can be activated
directly with the aid of IOT.
2. Hydrological and operational data are used to estimate opening and
closing of water gate that will ensure optimum supply of water.
4 Architecture
As discussed, the solution to our problem area is considered at three sections, i.e.,
dam authority, researcher and the common people/farmers. This way we create a synergy
between the stakeholders of the system.
Page 4

The architecture of the proposed solution is as per the figure given bellow;
Above architecture describes the system. The various components involved in the
system are:
• Ultrasonic Sensor
• Raspberry Pi
• Wi-Fi module
• Alarm system (siren)
• Dashboard
Integrate ultrasonic sensor and raspberry pi to collect the dam parameter (specifically
water level of the reservoir). This data along with the weather forecast from different
sources (like AccuWeather and meteorological department) are made accessible by
pushing the data to a cloud system. These two data (water level and weather forecast)
constitutes the hydrological data. Hydrological data along with operational data (data
based on experience and previous action) serve as an input to Intelligent Decision Support
System (IDSS). IDSS is a combination of DSS and artificial intelligence (AI). The
Page 5
Weather
forecast from
various
sources
Raspberry pi
Ultrasonic sensors
Dashboard and
central control unit
Alarm system

intelligent decision support model is based on Neural Network (NN), a mathematical
computational model that imitates the biological neuron capability. The theoretical
foundation and logic of NN has been discussed in [3] and [4]. The model consists of three
major stages:
o Data extraction
o Water level forecasting
o Water release decision modules
Page 6
Reservoir
Hydrological
Data
Operational
Data
Data MiningData Mining
Water Level
Forecasting
Model
Water Release
Decision Model
Water Release
Decision Model
Lock-
Gate
Opening
Lock-
Gate
Opening
Rainfall, Water
level
Experience and
Previous Action

Backpropagation algorithm will be used for learning and decision making of neural
network. Backpropagation algorithm looks for the minimum value of the error function in
weight space using a technique called the delta rule or gradient descent. The weights that
minimize the error function are then considered to be a solution to the learning problem
[9].
Backpropagation Algorithm:
initialize network weights (often small random values)
do
for each training example named ex
prediction = neural-net-output(network, ex) // forward pass
actual = teacher-output(ex)
compute error (prediction - actual) at the output units
compute {displaystyle Delta w_{h}} for all weights from hidden layer to output
layer // backward pass
compute {displaystyle Delta w_{i}} for all weights from input layer to hidden layer
// backward pass continued
update network weights // input layer not modified by error estimate
until all examples classified correctly or another stopping criterion satisfied
return the network
Page 7

The result based on NN model is sent to cloud and put to use under two different
situations.
Normal or regular situation:
The decision and time estimation of opening and closing the water gate is
provided to the dam authority via dashboard linked to cloud system. These data
will help the dam water gate operator to take a rational decision.
Critical or alarming situation:
A raspberry pi linked to the cloud system will trigger the alarm system (siren)
connected to it and thus informing the population residing in the flood prone area
about the emergency situation so that the region can be evacuated as soon as
possible.
5 Productization
The main purpose of our system is to assist the concerned dam authority with real-time
dam parameter and analytics using intelligent decision support system (IDSS) based on
Neural Network which has an accuracy rate of 96%. In practice, the water gate opening is
based on some operating rules; these rules do not consider the dynamic nature of
hydrology system. Therefore it is vital to use non-structural approach such as forecasting
to cope up with the event frequency and trigger alert to the authority when the situation is
severe. The proposed system is very flexible and easy to install. The system is divided
into six modules:
Page 8
CloudCloud
Flood
alert
Flood
alert
Raspberry piRaspberry pi
Alert messages for flood
using Buzzer

• Readings from sensors to raspberry pi.
• Data from Raspberry pi to cloud system.
• Weather forecasts from different sources to the cloud system.
• Processing of these data along with operational data (experience and
previous action) using Intelligent Decision Support System (IDSS).
• Analytics from IDSS to Dashboard via cloud system.
• Alarm system in case of emergency.
The above modules are integrated to form the smart system.
6 Tools and Environment
The following tools and environment will be used at various stages of our proposed
system.
6.1 Simulation & Testing
MATLAB Will be used for simulating the system. Test data for simulation will be
generated using MarkSim DSSAT weather file generator [5].
6.2 Cloud
We will be using ThingSpeak, an open-source Internet of Things (IOT)
application and API to store and retrieve data from things using the HTTP
protocol over the Internet or via a Local Area Network.
6.3 Physical
The following hardware, sensor and software component will be used in our
proposed system:
 Raspberry Pi: It is a small computer which can be programmed. In our
project we have used Raspberry Pi 3 model B. It is the latest product in the
Raspberry Pi 3 range. Specification has been mention in [6].
Page 9

 Ultrasonic Sensor: In our project we have used ultrasonic sensor (HC-
SR04) to determine the water level. It uses sonar to determine the water
level. Specification has been mention in [7].
 PyCharm: It is an integrated development environment (IDE) used
in computer programming, specifically for the Python language. We used
this platform to implement the Neural Network model.
6.4 Interface
Dashboard will be used for human interface. Dashboard can be easily created
using Bootstrap, Mysql and PHP.
7 References
[1] https://nptel.ac.in/courses/IIT-MADRAS/Hydraulics/pdfs/Unit41/41_2.pdf
[2] https://www.damsafety.in/ecm-includes/PDFs/DRIP_II_Presentation/Dam%20Safety
%20in%20India.pdf
[3] https://cdn.preterhuman.net/texts/science_and_technology/artificial_intelligence/Neur
al%20Networks%20-%20A%20Comprehensive%20Foundation%20-%20Simon
%20Haykin.pdf
[4] https://www.doc.ic.ac.uk/~nd/surprise_96/journal/vol4/cs11/report.html
[5] http://gismap.ciat.cgiar.org/MarkSimGCM/
[6] https://www.raspberrypi.org/products/raspberry-pi-3-model-b/
[7] https://www.tutorialspoint.com/arduino/arduino_ultrasonic_sensor.htm
[8] https://en.wikipedia.org/wiki/Backpropagation#/media/File:ArtificialNeuronModel_e
nglish.png
[9] https://www.edureka.co/blog/backpropagation/
Page 10

Smart Dam Monitering & Controling

More Related Content

What's hot

What's hot (20)

Similar to Smart Dam Monitering & Controling

Similar to Smart Dam Monitering & Controling (20)

Recently uploaded

Recently uploaded (20)

Smart Dam Monitering & Controling