Efficient Regional Traffic Signal Control Scheme Based on a SVR Traffic State Forecasting Model

Intelligent control of urban traffic signal takes a very important role in the Internet of Vehicle IOV. This paper proposes an intelligent control method to satisfy the real-time and accuracy of the regional traffic signal control ICMRT. On the basis of the existing wireless sensor network structure, ICMRT adopts the unequal clustering strategy to create a model of discrete switched system for the regional traffic system. Furthermore, taking the network delay and packet loss rate in data transmission into consideration, the state observer of the discrete switched system uses the improved $$\varepsilon $$-SVR theory to realize the online prediction of the multi-source data based traffic state, so that the overall traffic signal can be coordinately controlled. The asymptotic stability of discrete switched system is proved using Lyapunov function. Finally, simulation results show that ICMRT has better performance in the intersection average delay time compared with ordinary fuzzy neural prediction and ordinary $$\varepsilon $$-SVR method. So we can get the conclusion that this scheme is feasible and effective.