A Survey on Annual Average Daily Traffic (AADT) Driven Models for Traffic Forecasts

Abstract:- Off late, statistical and evolutionary algorithms are being extensively used for smart traffic systems whose sub-section is Intelligent Transportation Systems (ITS). For that purpose, different methods are being explored to estimate or forecast the traffic volume in a geographic are under different prevailing conditions so as to monitor and control massive traffic volumes, which has become a serious challenges in urban and even semi-urban areas worldwide. The prediction or forecasting problem is challenging since the nature of the data is extremely random and uncorrelated. A clear functional relationship in terms of correlation or regression analysis is seldom preset. Hence conventional statistical algorithms are being explored in the pretest which an adapt parameters as per the changing statistical properties of the fed data. This paper presents a comprehensive review on the need for evolutionary statistical algorithms for traffic forecasting problems and also cites the salient points of the existing literature. Moreover a comprehensive review of existing statistical algorithms used hitherto, are also cited. Finally the performance metrics are explained to evaluate the performance of such algorithms. This comprehensive review is expected to serve as a baseline for further research in the domain.