Multi-layer Fuzzy Control Model Research of Isolated Intersection

This paper proposed a multi-layer fuzzy control model for isolated intersection. The model mainly contains three layers: the first layer is prediction of traffic demand intensities, and it utilized a comprehensive index, Traffic Demand Intensities (TDI), to estimate demand degree of each approach lane to green time. The second layer is phase sequence optimization, which adjusts part of phase sequence on the base of certain phase sequence to adapt to the stochastic of traffic follow. The function of third layer is to optimize green time. With the input of traffic demand intensities and phase sequence optimization result, green time of every phase can be optimized. Compared with fixed-time control, vehicle-actuated control, this control model is effective in reducing average delay at intersection.