With the rapid growth in urban population and vehicle ownership, traffic congestion has become a severe problem everywhere in the world and is only expected to rise. This problem can be avoided by knowing the traffic situation in advance which is achieved with the help of traffic flow prediction. In the proposed work, traffic flow is predicted on short term basis using neuro-fuzzy hybrid system in combination with rough set. The neuro-fuzzy hybrid system combines the complementary capabilities of both neural networks and fuzzy logic. The work has attempted to study the effect of aggregation intervals and past samples on the prediction performance using MSE threshold variation. Rough set is used as a post processing tool. The objective is to improve prediction accuracy. Data from highway of Chennai, India is used for the analysis. It is found that use of rough set results in considerable improvement in the prediction performance as indicated by performance measures like MSE, RMSE etc.
Minal Deshpande
GH Raisoni College of Engineering, India
Intelligent Transportation Systems (ITS), Rough Set Theory (RST), Short Term Traffic Flow Prediction, Neuro-fuzzy Hybrid System | Published By :
ICTACT
Published In :
ICTACT Journal on Soft Computing
( Volume: 10 , Issue: 3 )
Date of Publication :
April 2020
|