An intelligent transportation system (ITS) is a major pillar in the development of smart cities. Real-time short-term traffic flow forecasting models are vital in the implementation of ITSs. The GM(1,1) is one of the prediction models which has been employed before in forecasting time series events and in this paper we improve the precision of the original, GM(1,1), by combination of a data grouping technique (DGT) and modification of its background value (MBV). We establish an improved grey model denoted by MBVGGM(1,1). In addition we perform short-term traffic flow forecasting by the improved GM(1,1) as an important element in developing ITSs. The results show that the DGT significantly improves the accuracy of the grey model in fitting and forecasting of traffic flow. Moreover, combination of DGT and MBV methods greatly improves short-term forecasting accuracy compared with the original GM(1,1). Thus the new knowledge in this paper will enhance transportation systems in major cities by improving their short-term traffic flow forecasting. Moreover, proactive vehicle flow control will easy traffic management systems on our roadways through ITSs.
Improved Grey Model, GM (1, 1), in Short-Term Traffic Flow Forecasting - Smart Transportation Systems
2019 IEEE AFRICON ; 1-9
2019-09-01
2884692 byte
Conference paper
Electronic Resource
English
Bayesian Time-Series Model for Short-Term Traffic Flow Forecasting
British Library Online Contents | 2007
|The Improved Forecasting Model for Short-Term Traffic Flow Based on Phase Space Reconstruction
British Library Conference Proceedings | 2009
|