50120140504004

International Journal of Computer Engineering and Technology (IJCET), ISSN 0976-6367(Print),

ISSN 0976 - 6375(Online), Volume 5, Issue 4, April (2014), pp. 32-40 © IAEME

32

AN INTELLIGENT FUZZY-BASED TSUNAMI WARNING SYSTEM

Twinkle Tayal1, Dr. Prema K.V.

2

1(2

nd year, M.Tech, CSE, FET, MUST, Rajasthan, India) 2(Dept of CSE, FET, MUST, Rajasthan, India)

ABSTRACT

A tsunami is known as progression of water waves prompted by the dislodgement of a

substantial volume of a body of water, generally an ocean or a large lake. There are various factors

that can generate a tsunami like Earthquakes, volcanic eruptions and other underwater explosions,

landslides, glacier calvings, meteorite impacts and other disarticulations above or below water.

Tsunamis obliterate not only human population but all other species. There are number of

confident ways to envisage such disasters and design diverse kinds of early warning systems.

These can be prophesized on the climatic conditions and several other parameters. With the

overture of modern science and computer technology, the field of Artificial Intelligence is showing

an explicit utility in all spectrums of life. One such concept that is functioning as a detonation in

the fields of environmental science and policy is fuzzy logic. In this work, we will try to foretell

the tsunami based on the certain factors. All the parameters taken for this work are real- time and

the data used is collected from the well-known organizations such as NOAA pacific tsunami

warning centre, Japan meteorological agency, UNESCO international tsunami information centre.

The system has been designed in the Matlab Fuzzy Logic Toolbox. The system designed by us is

also compared with an existing system in this paper.

Keywords: Tsunami, Tsunami Prediction, Tsunami Warning, Fuzzy, Fuzzy Logic.

1. INTRODUCTION

Tsunamis are among the most detrimental natural disasters known to man. For most of the

people who live close to sea shore, tsunami is the greatest menace of their live. Tsunami causes

rivers and other water paths to brim over. This superfluous water can generate noxious currents and

drag away people, causing them to drown. A tsunami has all of the destructive effects plus the

added destructive power crashing waves [1]. As shown in fig.1, most of the oceanic tsunamis (up

to 75% of all historical cases) are triggered by shallow-focus earthquakes dexterous of transferring

INTERNATIONAL JOURNAL OF COMPUTER ENGINEERING &

TECHNOLOGY (IJCET)

ISSN 0976 – 6367(Print)

ISSN 0976 – 6375(Online)

Volume 5, Issue 4, April (2014), pp. 32-40

© IAEME: www.iaeme.com/ijcet.asp

Journal Impact Factor (2014): 8.5328 (Calculated by GISI)

www.jifactor.com

IJCET

© I A E M E

International Journal of Computer Engineering and Technology (IJCET), ISSN 0976

ISSN 0976 - 6375(Online), Volume 5, Issue 4, April (2014), pp.

enough energy to the overlying water colu

volcanic (5%) and meteorological (2%) tsunamis.

clandestine sources. Though the effect of tsunamis is

disastrous power can be colossal

numerous tsunamis recorded in the history that were devastating and were very

necessitate designing such methods

tsunami.

Figure

The increasing esteem of artific

fuzzy logic as a method to solve this problem and carry out this work

utilized for foretelling tsunami because of the

the nature of tsunami and different source

tsunami. So, in this paper, we are proposing a fuzz

tsunami based on the different parameters. All the data regarding the parameters are collected from

the historical databases provided by the

Japan meteorological agency, UNESCO internati

2. FUZZY LOGIC SYSTEMS

A. Lotfi Zadeh, a professor at the University of California

logic at Berkley. He presented this

processing data by allowing fractional or partial set membership rather than

or non-membership. Fuzzy logic means

with words and so on. It bestows mathematical strength to the emulation of specific per

linguistic traits associated with human cognition

form of verbal phrases or linguistic terms suc

values. If a system’s behaviour can be

processes, fuzzy logic approach can be

technique to a real application requires the following thre

1. Fuzzification – it converts classical data or crisp data into fuzzy data or Membership

Functions (MFs).

2. Fuzzy Inference Process – it coalesce

the fuzzy output.

3. Defuzzification – it uses several

International Journal of Computer Engineering and Technology (IJCET), ISSN 0976

6375(Online), Volume 5, Issue 4, April (2014), pp. 32-40 © IAEME

33

ying water column. The rest is estranged between the landslide (7%),

volcanic (5%) and meteorological (2%) tsunamis. Up to 10% of all the historical run

Though the effect of tsunamis is constrained to the coastal areas, their

colossal and they can influence the entire ocean basins. There are

sunamis recorded in the history that were devastating and were very

such methods that can be used to warn people beforehand

Figure -1: causes of tsunami [1]

of artificial intelligence in the numerous fields, make us to

method to solve this problem and carry out this work. In this work

because of the undeniable reason that there is a natural

e of tsunami and different sources of tsunami persuade differently on the occurrence of

, we are proposing a fuzzy expert system that will alert about the


cal databases provided by the organizations like NOAA pacific tsunami warning centre,

meteorological agency, UNESCO international tsunami information centre.

A. Lotfi Zadeh, a professor at the University of California, introduced the concept of fuzzy

at Berkley. He presented this notion not as a control methodology, but as a

fractional or partial set membership rather than crisp set

Fuzzy logic means inexact reasoning, information granulation, computing

mathematical strength to the emulation of specific per

s associated with human cognition [2]. In fuzzy logic, information is

linguistic terms such as big, small, very, few etc despite of

values. If a system’s behaviour can be uttered by rules or entails very complex non

processes, fuzzy logic approach can be useful in that system [3]. To put into practice,

technique to a real application requires the following three steps:

classical data or crisp data into fuzzy data or Membership

coalesce membership functions with the control rules to

s several methods to calculate each allied output [4].


between the landslide (7%),

historical run-ups still have

to the coastal areas, their

ocean basins. There are

sunamis recorded in the history that were devastating and were very damaging. It is

to warn people beforehand regarding the

fields, make us to prefer

In this work, Fuzzy Logic is

reason that there is a natural uncertainty in

persuade differently on the occurrence of

y expert system that will alert about the incident of


NOAA pacific tsunami warning centre,

, introduced the concept of fuzzy

methodology, but as an approach of

crisp set membership

reasoning, information granulation, computing

mathematical strength to the emulation of specific perceptual and

In fuzzy logic, information is presented in

h as big, small, very, few etc despite of numeric

very complex non-linear

put into practice, fuzzy logic

classical data or crisp data into fuzzy data or Membership

control rules to derive



34

3. PRELIMINARIES

It was a challenge for us to perceive the pertinent parameters in the beginning. So, a vigilant

study was done and the best and most useful possible set of input parameters was selected, that

could be ample to model the real time scenario and help us to precisely predict and warn about the

tsunami. In total, 5 inputs are used in this fuzzy system.

1. Earthquake (EQ) – it is the most significant cause of tsunamis. Different scales are used for

measuring the earthquakes. Here, measurements in Richter scale are used. Generally, it has

been observed that earthquakes above 6.5 are supposed to cause tsunamis.

2. Focal depth (FD) - It is depth of an earthquake hypocenter (the point within the earth where an

earthquake shatter starts). Tsunami when caused by earthquake, it also depends on the focal

depth. The shallow focal earthquakes are most destructive. We have taken 0 to 65 km as range

for shallow focal depth in this work.

3. Volcanic eruption index (VEI) - on land eruptions or underwater volcanic eruptions can cause

tsunamis. VEI (volcanic eruption index) is used to define the kinds of eruptions as explosive or

not. Optimal range of VEI is from 0 to 8.

4. Landslide (LS) - Landslides stepping into oceans, bays, or lakes can also cause tsunami.

Generally, such landslides are generated by earthquakes or volcanic eruptions.

5. Height of waves in deep ocean (WD) – in deep ocean height of tsunami wave is very less, not

more than 4meters due to decreased level of potential energy. As waves reach at the shore, the

height of waves keeps on increasing.

4. METHODOLOGY

Fuzzy inference system for tsunami warning system can be designed by applying following

procedure in the Matlab Fuzzy Logic Toolbox:

1. Look over the problem to be solved and decide the input and output variables.

2. Deciding the fuzzy inference rules. This usually depends on human familiarity, understanding

and trial-and-error.

3. Fuzzy membership functions for all the inputs and the output. Fuzziness in a fuzzy set is

illustrated by its membership functions. It recognizes the element in the set, if it is discrete or

continuous.

4. Perform fuzzy inference based on the inference method. Smoothness of the final control

surface is resolute by the inference and defuzzification methods.

5. Select a defuzzification method. Defuzzification means the conversion of fuzzy to crisp.

In this work, we are using the Mamdani method for the compelling reasons as it is spontaneous,

commonly used, extensively accepted and it is suited to system requiring human intervention. In the

present work, system is developed by using the GUI tools, which consists of five editors to build,

edit and view the system, as shown in fig.2, namely

1. Fuzzy Inference System (FIS) Editor – utilized for handling the issues for the system like

number of input and output variables and their names.

2. Membership Function Editor- used for defining shapes of all the membership functions allied

with each variable.

3. Rule Editor- utilized to edit the list of rules that defines the behavior of the system.



35

4. Rule Viewer- used to view the fuzzy inference diagram. It is used to see which rules are active,

or how individual membership function shapes influence the results.

5. Surface Viewer – it is utilized to view the dependency of one of the outputs on any one or two

of the inputs. It generates and plots an output surface map for the system.

Precisely, a fuzzy decision is the upshot of weighing the facts and its significance in the same

way as humans take decisions. Fuzzy logic replicates human like thinking where the human can

figure out a vague inference from an assortment of imprecise premises [5].

Figure – 2: GUI editors in Mamdani fuzzy method [5]

The overall fuzzy inference model for tsunami prediction system can be shown as in the fig.3.

Figure -3: fuzzy inference system for tsunami prediction

4.1. Input/ Output Membership Functions There are 5 inputs in this system. Each input is defined by using the different

membership functions. The output alert is described by the membership functions rare, advisory



36

and warning. The output of system will be rare if no tsunami is likely to occur, advisory if there

may be chances of tsunami in near future and warning if tsunami is definite and can be destructive.

These functions symbolize a degree of a binary value, 1 being the highest and 0 being the lowest.

All the inputs and output are described by trapezoidal membership function to maintain uniformity

in the system. All the membership functions are shown in the Table 1 and the snapshots of Matlab

Fuzzy Logic Toolbox in the figures below.

TABLE 1: Membership Functions of Inputs and Output

Figure -4: membership function for input EQ

Figure-5: Membership Function for input VEI.

Variable Membership functions

EQ {WEAK,MILD,STRONG}

FD {SHALLOW,MODERATE,DEEP}

VEI {NON_EXPLOSIVE,MILD,EXPLOSIVE}

LS {WEAK,MILD,STRONG}

WD {TSUNAMI,MAYBE,NORMAL}

ALERT(OUTPUT) {RARE, ADVISORY, WARNING}



37

Figure- 6: Membership Function for input LS

.

Figure – 7: Membership Function for input FD

Figure -8: Membership Function for input WD



38

Figure – 9: Membership Function for output ALERT

4.2. Fuzzy Rules Fuzzy rules are the most imperative part in the fuzzy system. These rules can be selected on

the basis of one’s knowledge, perception or understanding of the problem. As here, we are using the

standardized data for all the inputs; fuzzy rules are dependent on them. The fuzzy rules are in the

form of IF x then y. The rules in our system are defined in the following way:

1. If (EQ is WEAK) and (VEI is MILD) and (LS is MILD) and (FD is SHALLOW) and (WD is

MAYBE) then (ALERT is WARNING)

2. If (EQ is STRONG) and (VEI is MILD) and (LS is MILD) and (FD is MODERATE) and (WD

is MAYBE) then (ALERT is ADVISORY)

3. If (EQ is MILD) and (VEI is NON_EXPLOSIVE) and (LS is MILD) and (FD is DEEP) and

(WD is NORMAL) then (ALERT is RARE)

The rule editor for the system of Matlab Fuzzy Logic Toolbox is shown in figure 9.

Figure – 10: Rule Editor



39

4.3. Simulations We have accomplished a number of experiments by entering the different values of the inputs

and every time the system is giving correct output as it was supposed according to our perception

and information according to historical databases. Rule viewer can be used to enter the inputs and

see how each and every rule is behaving on your given input. Finally, it gives one defuzzified crisp

output based on the method you had used. When the decisive parameters are in range of warning, the

input values in this situation for the different parameters are [9 0 0 10 2] and the corresponding rule

viewer is shown in fig. 11.

Figure – 11: rule viewer when decisive parameters are in warning range

The result for the input given above is coming out to be 0.8276, which is under the warning

range. When the output alert should be advisory: when we input values as [9 0 0 10 9], the

defuzzified value comes out to be 0.5, which is under advisory range. When the output alert should

be rare: rare will be the alert when no risk of tsunami is there. When we input values as [4 0 0 300

9], the defuzzified value comes out to be 0.169, which is under rare range.

5. RESULTS AND ANALYSIS

On carrying out a number of experiments with different data sets, we are getting the correct

output every time as supposed. We have compared our work with the work in the IEEE research

paper “Cherian, Carathedathu Mathew, Nivethitha Jayaraj, and S. Ganesh Vaidyanathan,

Artificially Intelligent Tsunami Early Warning System, 12th International Conference on

Computer Modeling and Simulation (UKSim), 2010, pp. 39-44. IEEE, 2010”[6]. In this paper,

they have taken only 2 parameters, but we have taken 5 inputs as according to our detailed study of

NOAA tsunami historical database [7] that consists information on tsunami events from 2000 B.C. to

the present in the Atlantic, Indian, and Pacific Oceans; and the Mediterranean and Caribbean Seas,

along with the earthquake, landslides and volcanic eruptions are also causes tsunami and tsunami

largely depends on the focal depth. They had also implemented the problem in the Matlab Fuzzy

Logic Toolbox as we have. We have taken standardized and real- time data according to the well

known authorities NOAA pacific tsunami warning centre, Japan meteorological agency, UNESCO

international tsunami information centre and defined our membership functions for each input

according to the standard ranges. As they had taken 2 inputs, they are using 12 rules in their system

whereas we are using 159 rules to describe the system,. We have uniformity in our system as we



40

have described all the inputs by the trapezoidal membership functions. They had divided the

different parameters into various no. of ranges, like for output they are using 5 sets, where as we are

taking 3 sets namely rare, advisory and warning, which is a more realistic situation and simple to

recognize. Moreover, when compared with this existing system, our system is found to be more

capable and is giving better outcomes.

6. CONCLUSION

Fuzzy logic imparts a complementary approach to signify linguistic and subjective facets of

the real world in computing. The intent behind choosing fuzzy logic in this work is that system

uses fuzzy logic model put across valuable and real results depending on the uncertain, vague,

inconclusive, indecisive and imprecise verbal acquaintance just like logic of a human being.

Moreover, it takes long time to use other existing methods for such problems whereas we can reach

a general solution by doing only limited number of experiments in fuzzy. Mamdani has been

designed in this study. The prediction scheme presented here can be deliberated as a step towards

the prediction of this destructive natural hazard tsunami. This can successfully be pertained by

taking other parameters into consideration and moreover, in this study, we have used general data

evaluated from the historical database, this system can work more efficiently if data for a particular

area is used.

7. REFERENCES

[1] Sidharth Das, BiramBaburayBaskey, Design of an Embedded System for the Detection of

Tsunami, B.tech Dissertation, Dept. Electronics and communication Eng., National Institute of

Technology, Rourkela, May, 2012.

[2] Ying Bai and Dali Wang, Fundamentals of Fuzzy Logic Control – Fuzzy Sets, Fuzzy Rules

and Defuzzifications, Advanced Fuzzy Logic Technologies in Industrial Applications,

Springer, 2006.

[3] Poongodi, M., Manjula, L., Pradeepkumar, S. and Umadevi, M, Cancer prediction technique

using fuzzy logic, International journal of Current Research, Vol. 3, Issue 11, pp. 333-336,

Dec., 2011.

[4] Lotfi a. Zadeh, Knowledge Representation in Fuzzy Logic, IEEE TRANSACTIONS ON

KNOWLEDGE AND DATA ENGINEERING, VOL. I, NO. I, MARCH 1989.

[5] Sivanandam, S. N., Sai Sumathi, and S. N. Deepa. Introduction to fuzzy logic using MATLAB.

Vol.1. (Berlin: Springer, 2007).

[6] Cherian, Carathedathu Mathew, Nivethitha Jayaraj, and S. Ganesh Vaidyanathan, Artificially

Intelligent Tsunami Early Warning System, 12th International Conference on Computer

Modeling and Simulation (UKSim), 2010, pp. 39-44. IEEE, 2010”

[7] National Geophysical Data Center / World Data Service (NGDC/WDS): Global Historical

Tsunami Database. National Geophysical Data Center, NOAA, doi: 10.7289/V5PN93H7.

[8] Mohammed Sirajuddin, Dr D. Rajya Lakshmi, Dr Syed Abdul Sattar and Nafisur Rahman,

“Fuzzy Logic — The Fascinating Logic Behind Artificial Computational Intelligence”

International Journal of Advanced Research in Engineering & Technology (IJARET),

Volume 4, Issue 3, 2013, pp. 280 - 285, ISSN Print: 0976-6480, ISSN Online: 0976-6499.

50120140504004

Documents