A real-time recurrent learning on predicting short-term temporal traffic dynamics for sustainable management
Short-term prediction of traffic dynamics to mitigating congestions remains critical in the field of sustainable transport systems. In this paper, a real-time recurrent learning algorithm (RTRL) is proposed to address the above issue. Furthermore, the authors also dabble in comparing pair predictability of linear method vs. RTRL algorithms and simple non-linear method vs. RTRL algorithms using a first order autoregressive time-series AR(1) and a deterministic function. A field study tested with flow, speed and occupancy series data collected from dual-loop detectors on a freeway is conducted. The numerical results reveal that the performance of RTRL on predicting short-term traffic dynamics is satisfactorily accepted. Furthermore, it is found that the dynamics of short-term traffic states characterised in different time intervals, collected in diverse time lags and times of day may have significant effects on the prediction accuracy of the proposed algorithms. Such findings have efficiently raised the performance for sustainable transport systems.
Keywords: real-time recurrent learning, RTRL, traffic dynamics, congestion management, sustainable transport, sustainability, sustainable development, short-term prediction