Majalah Ilmiah UNIKOM

Vol.8, No. 2

214

H a l a m a n

Muhammad Aria

6 show the comparison of T2FL, T1FL and

fixed-time computation time. So Interval

Type 2 TSK Fuzzy system takes 3.6 times

slower than Fixed Timer and 1.8 slower than

Type 1 TSK Fuzzy algorithm. In application,

Interval Type 2 TSK Fuzzy algorithm needs

17888 units of memory, while Fixed Timer

only needs 4840 unit memory and Type 1

TSK Fuzzy algorithm just needs 13032 units

memory.

CONCLUSION

In this paper, we have proposed the traffic

controller for complex intersection group

based on interval type 2 TSK fuzzy systems

and implement the simulator for perform-

ance evaluations. To control a set of inter-

section, we distribute controls to each con-

troller. Each controller takes charge of con-

trolling its traffic signal and cooperating with

its neighborhood. Our approach can be eas-

ily extended to any situation. According to

the simulation studies, results of the fuzzy

logic controller is the same as the fixed-time

controller in normal traffic flow. But the

simulation shows promising results in the

cases of heavy traffic and time-varying traf-

fic with large variance. In the cases of heavy

traffic and time-varying traffic, T2FL algo-

rithm reduced average vehicle delay 13,2 %

better than Fixed Timer and 1,6 % than T1FL

algorithm. But T2FL computation time is

more complex than both Fixed Timer and

T1FL algorithm, so T2FL system takes 3.6

times slower than Fixed Timer and 1.8

slower than T1FL algorithm. In application,

T2FL needs 17888 units of memory, while

Fixed Timer only needs 4840 unit memory

and T1FL just needs 13032 units memory.

REFERENCES

Mohammad Hossein Fazel Zarandi, Shab-

A Fuzzy Signal Controller

for Isolated Intersections

Uncertain System, vol 3, No. 3, pp 174

– 182, 2009

Lin Zhang, Honglong Li, Panos D. Preve-

Signal Control for Oversaturated

Intersections Using Fuzzy Logic

portation Research Record, Hawaii,

2004

Jee-Hyong Lee, Keon-Myung Lee, KyoungA

Seoing, Chang Bum Kim and Hyung Lee-

Traffic Control of Intersection

Group Based on Fuzzy Logic

Marzuki Khalid, See Chin Liang and Rubiyah

Control of a Complex Traffic Jun-

cion using Fuzzy Inference

Tan Kok Khiang, Marzuki Khalid and Rubi-

Intelligent Traffic Lights Con-

trol by Fuzzy Logic

A New

Approach for Fuzzy Traffic Signal Con-

trol

Type-2 Fuzzy Sets

and Systems : An Overview

putational Intelligence Magazine, vol 2,

no.1, pp. 20-29

Ching-Hung Lee, Yu-Ching Lin, and Wei-Yu

Systems Identification Using Type-2

Fuzzy Neural Network (Type-2 FNN) Sys-

tems

Symposium on Computational Intelli-

gence in Robotics and Automation,

2003

Adaptive

Noise Cancellation Using Type-2 Fuzzy

Logic and Neural Networks

Fuzzy 2004, IEEE press, 2004

Case

Fixed-Time

T1FL

T2FL

2.a

57.0

47.8

48.3

2.b

61.1

56.9

53.0

2.c

110.9

72.6

74.7

Table 5. Average delay time for Case 2

Average of time

computation

(microsecond)

Fixed-Time

9375

T1FL

18750

T2FL

33854

Table 6. Computation time