Prepared By
A Project Report
To be submitted in the partial fulfillment of the requirements
For the degree of
Affiliated to
CERTIFICATE
Countersigned By
_______________
_______________
_______________
_______________
_______________
iii
ACKNOWLEDGEMENT
iv
LIST OF FIGURES
v
LIST OF TABLES
Table 7.2: Input & Output values based on real world test values. 28
vi
ABSTRACT
The detection and prevention of forest fire is a major problem now a day.
Timely detection allows the prevention units to reach the fire in its initial
stage and thus reduce the risk of spreading and the harmful impact on
human and animal life. Because of the inadequacy of conventional forest
fire detection in real time and monitoring accuracy the Wireless Sensor
Network (WSN) is introduced. This project proposes a fuzzy logic based
implementation to manage the uncertainty in forest fire detection
problem. Sensor nodes are used for detecting probability of fire with
variations during different time in a day. The Sensor nodes sense
temperature, humidity, light intensity, CO2 density and time and send the
information to the base station. This proposed system improves the
accuracy of the forest fire detection and also provides a real time based
detection system as all the input variables are collected in real time basis.
vii
TABLE OF CONTENTS
Certificate ii
Acknowledgement iv
List of Figures v
List of Tables vi
Abstract vii
Chapter 1 Introduction 1
Conclusion 37
Bibliography 38
FUZZY LOGIC BASED IMPLEMENTATION FOR FOREST FIRE
DETECTION USING WIRELESS SENSOR NETWORK
INTRODUCTION
1
CHAPTER 1
INTRODUCTION
1
alarming rate. In order to minimize damage, early detection of forest fires
is a crucial issue. Without a clear and correct understanding of the
distribution and dynamics of forest fires, it is impossible to effectively
manage them. Thanks to the technology that enables the deployment of
devices called motes, in large numbers directly into fire zones, wireless
sensor networks (WSNs) can significantly improve the accuracy and
density of parametric measurement of physical phenomena. It will be
beneficial to detect the pre-cursors of these disasters, early warn the
population, evacuate them, and save their life.
2
conventional forest fire detection on real time and monitoring accuracy the
Wireless Sensor Network (WSN) is introduced.
3
FUZZY LOGIC BASED IMPLEMENTATION FOR FOREST FIRE
DETECTION USING WIRELESS SENSOR NETWORK
RELATED WORKS
2
CHAPTER 2
RELATED WORKS
5
the conventional WSN detection approaches by reducing the number of
false alarms, authors proposed an image-based real time fire detection
technique. Unfortunately, most of these studies choose simulating their
proposed solutions instead of doing experiments in real test-bed
environments, since that kind of setup exposes additional difficulties. Even
those using test-bed to carry out real experiments; they have not made a
serious study on which detection methods could be very suitable to their
context. In the context of the above studies, we propose a comparative
study between two forest fires detection methods (Canadian and Korean)
using a real test-bed based approach to choose the one that fits the context
of our country.
Canadian approach
The Canadian study proposed the calculation of the index fire according to
FWI (Fire Weather Index). This eliminates the need to communicate all
the sensor data to Sink, and only a few aggregated index are reported for
reduce energy consumption. FWI system comprises six standardized
index. The three first shows daily variations of water content of three types
of fuel forest with different speeds drying. The other three relate to fire
behavior and are representative of the propagation speed, the quantity of
burned fuel and intensity of the fire. The method is based solely on the
determination noon daily weather: temperature, relative humidity, speed
wind and rain during the last 24 hours (if there was). The month must also
be specified. This method is primarily to solve a set of equations (Van
Wagner and Pickett, 1985), which can be calculated with fast computer.
Korean approach
This approach is implemented on the system FFSS (Forest-fires
Surveillance System). The middleware developed in this study receives and
6
processes packets from the transceiver and displays its results. The results
contain the level of risk of forest fires. This level is calculated by the
formula defined by the following equation:
P is rain (mm).
Then, the software saves the received packets to database server and
generates emergency alerts by the Note that other systems for detecting
forest fires can be found in practice such as National Fire Danger Rating
System (NFDRS) and D-FLER (Distributed Fuzzy Logic Engine Rule-based
WSNs).
7
FUZZY LOGIC BASED IMPLEMENTATION FOR FOREST FIRE
DETECTION USING WIRELESS SENSOR NETWORK
PROBLEM DEFINITION
&
OBJECTIVES
3
CHAPTER 3
PROBLEM DEFINITION & OBJECTIVES
9
In computer science and telecommunications, wireless sensor networks
are an active research area with numerous workshops and conferences
arranged each year, for example IPSN, SenSys, and EWSN.
10
FUZZY LOGIC BASED IMPLEMENTATION FOR FOREST FIRE
DETECTION USING WIRELESS SENSOR NETWORK
4
CHAPTER 4
PROPOSED SOLUTION FOR FOREST FIRE
DETECTION
4.1 Fuzzification
Fuzzy Logic was introduced in 1965 by Lotfi A. Zadeh, who was professor
in computer science at the University of California in Berkeley. Fuzzy Logic
is a multi-valued logic that allows intermediate values to be defined
between conventional evaluations like true/false, yes/no, high/low, etc.
12
Fuzzy Logic has emerged as a profitable tool for the controlling and
steering of systems, complex industrial processes, household,
entertainment electronics, as well as for other expert systems and
applications. The aim is to use fuzzy sets in order to make computers more
intelligent, therefore. Fuzziness describes event ambiguity and
impreciseness of linguistic terms. Fuzzy logic fits best in applications
where the variables are continuous and/or mathematical models do not
exist or traditional system models become overly difficult. WSN is typically
used to supervise some parameters of an environment process. The
atmospheric events are multifaceted, confusing and imprecision embedded
in their nature. Consequently, a fuzzy based approach is a feasible option.
The model of fuzzy logic system consists of fuzzification, fuzzy rules, fuzzy
inference system and de-fuzzification process. The Fuzzification is the first
step in the fuzzy inferencing process. This involves a domain
transformation where crisp inputs are transformed into fuzzy inputs. Crisp
inputs are the exact inputs measured by sensors and passed into the
control system for processing, such as temperature, position, pressure,
rpm's, etc. Each crisp input that is to be processed by the Fuzzy
Inferencing Unit has its own group of membership functions or sets to
which they are transformed. The group of membership functions exists
within a universe of discourse that holds all relevant values that the crisp
input can possess. In our fire detection algorithm temperature, relative
humidity, CO2 density and time act as input fuzzy variables. The
Probability of Fire is the output variable. The membership functions LOW,
MEDIUM and HIGH are defined on temperature, light intensity, humidity
and CO density whereas BEFORE NOON, NOON, AFTER NOON for.
VERY LOW, LOW, MEDIUM, HIGH and VERY HIGH are defined on
Probability of fire as shown in the table on the next page.
13
Membership
Variable Range
Functions
Membership
Variable Range
Functions
Legend:
VL: Very Low, L: Low,
N: Normal, M: Medium, O: Optimum,
H: High, VH: Very High,
EM: Early Morning, M: Morning, F: Forenoon, AF: Afternoon,
E: Evening, N: Night
14
FUZZY LOGIC BASED IMPLEMENTATION FOR FOREST FIRE
DETECTION USING WIRELESS SENSOR NETWORK
5
CHAPTER 5
Software Requirements:
Operating System: Windows XP or, Windows 7
Programming IDE: MATLAB 2013
Hardware Components:
Processor: Dual core processor. Preferably, Intel i3.
Hard Drive: A minimum of 10 GB space is required for the
software to run without problems.
Memory: A minimum of 4 GB RAM is required for faster
processing.
A sufficiently fast and reliable internet connection.
A wireless network of connected sensors for data accumulation.
16
FUZZY LOGIC BASED IMPLEMENTATION FOR FOREST FIRE
DETECTION USING WIRELESS SENSOR NETWORK
PROJECT IMPLEMENTATION
6
CHAPTER 6
PROJECT IMPLEMENTATION
6.1 Implementation
The toolbox lets us model complex system behaviors using simple logic
rules, and then implement these rules in a fuzzy inference system. We can
use it as a stand-alone fuzzy inference engine. Alternatively, we can use
fuzzy inference blocks in Simulink and simulate the fuzzy systems within a
comprehensive model of the entire dynamic system.
18
6.2 Rules for Evaluation
19
Fig 6.2: Screenshot of Rules 2
20
6.3 Input and Output Parameters
In the above image, the left side indicates the input parameters and the
right side indicates the output parameter. Various values of Temperature,
Relative Humidity, CO2 and Time are input on the left side. The middle
portion of the image indicates the Fuzzy Logic Toolbox which evaluates the
input to give the output on the right side, which is Fire Probability.
21
6.4 Range Of Values for Input Parameters
22
Fig 6.6: CO2 density range graph
23
6.5 Range Of Values for Output Parameter
24
6.6 Testing Of Rules for Output
25
FUZZY LOGIC BASED IMPLEMENTATION FOR FOREST FIRE
DETECTION USING WIRELESS SENSOR NETWORK
7
CHAPTER 7
USER INPUT AND OUTPUT
27
RESULT
Relative CO2 Time
Test Temperature Fire
Humidity Density (in 24 hour
Case (oC) format) Probability
(%) (ppm)
1 18 70 1000 22 10
2 18 50 1000 22 10
3 18 30 1000 22 10
4 18 30 2000 22 10
5 30 7 4000 14 50
6 30 60 1600 10 10.00
7 30 60 3000 14 28.43
8 30 50 4800 10 58.31
9 30 30 4800 14 68.31
10 30 14 4800 14 68.31
11 100 60 3000 10 48.43
12 100 40 3000 10 48.43
13 100 30 3000 10 78.43
14 100 10 4500 10 88.43
15 100 10 4500 13 88.43
16 100 60 3000 10 48.43
17 100 20 3000 13 78.43
18 500 10 4000 10 88.42
19 500 10 4000 10 88.42
20 500 10 4000 13 88.42
21 500 40 4000 13 88.42
22 500 20 4000 10 88.42
23 500 20 4000 13 88.42
24 500 40 4000 10 88.42
25 500 40 4000 13 88.42
Table 7.2: Input & Output values based on real world test values.
Based on the input values: Temperature, Relative Humidity, CO2
Density and Time, the output received is the Fire Probability.
Example:
Relative Time
Temperature CO2 Density
Humidity (in 24 hour Fire Probability
(oC) (ppm)
(%) format)
18 70 1000 22 10
28
FUZZY LOGIC BASED IMPLEMENTATION FOR FOREST FIRE
DETECTION USING WIRELESS SENSOR NETWORK
8
CHAPTER 8
GRAPHICAL VIEW OF RULES WITH INPUT AND
OUTPUT
30
Fig 8.2: Screenshot of Rule Viewer 2
31
Fig 8.3: Screenshot of Rule Viewer 3
32
FUZZY LOGIC BASED IMPLEMENTATION FOR FOREST FIRE
DETECTION USING WIRELESS SENSOR NETWORK
9
CHAPTER 9
PROJECT PLANNING AND SCHEDULING
34
Fig 9.2: Screenshot of Gantt Chart
35
FUZZY LOGIC BASED IMPLEMENTATION FOR FOREST FIRE
DETECTION USING WIRELESS SENSOR NETWORK
10
CHAPTER 10
FUTURE SCOPE AND FURTHER ENHANCEMENT
One area which can see lots of improvement is data gathering and
collection. Hardware that makes use of this program for forest fire
detection can be designed to operate based on natural triggers.
Forest fires, man-made or natural, are such catastrophes which boggle the
mind! Controlling, or even predicting forest fires is an immensely difficult
task. This is mostly due to presence of such a large number of parameters
that may speed up the fire, change the direction of the fire, etc. The
temperatures of the local area during a forest fire are so high that there is a
good chance of multiple sensors being damaged or being burnt out!
Therefore, one area of improvement can be the development of fire
resistant and robust hardware.
The software program can also be so implemented that it can take more
number of variables as input, with much vague values, and give very
accurate results.
Energy efficient and versatile hardware can also be very useful for proper
implementation of this project.
37
CONCLUSION
38
BIBLIOGRAPHY
6. A.K. Singh, and Harshit Singh, Forest Fire Detection through Wireless
Sensor Network using Type-2 Fuzzy System, International Journal of
Computer Applications, vol. 52 No.9, pp. 19-23,August 2012.
39