An algorithm for freeway traffic state detection considering speed difference characteristic

Traditional algorithms for freeway traffic state detection are based on macroscopic traffic flow parameters (i.e. average speed, volume and density). These macroscopic parameters are defined by accumulation or mean values in detection period and could not effectively describe the state difference among different vehicle. This paper proposes two speed difference parameters to study the speed difference characteristic. Then the freeway state detection algorithm for estimating real-time state is designed by the use of fuzzy c-means clustering model. Simulation experiments are performed to evaluate the performance of the algorithm in the non-current congestion and current congestion environment. The results show that the detected state can reflect the development trend of the vehicle queue length well and are consistent with the traffic simulation condition. Compared with the contrast algorithm, the algorithm has advantage in stability and accuracy, and the algorithm may be used as a promising method in freeway traffic state detection.