Online Coordination of Signals for Heterogeneous Traffic Using Stop Line Detection

Real time control of heterogeneous traffic is always a challenge for efficient and effective traffic management. The solution becomes more complex when the heterogeneity is augmented with limited lane discipline. Adaptive traffic signal controllers offer better signal time management especially when the traffic is not saturated. Traditional methods of upstream vehicle detection (UTOPIA, PRODYN, SCOOT, OPAC, and RHODES) work well with disciplined traffic. However, this approach has several limitations when the drivers do not adhere to their respective lanes. To address this, a traffic adaptive control model which uses stop-line detector information is proposed in this paper. In a corridor of closely spaced controlled intersection, vehicular movement can be improved largely by synchronizing the signals between intersections. Further, efficiency of the synchronization can be enhanced by operating the corridor at an optimum cycle. The model aims at real-time allocation of optimum cycle time through actor-critic reinforcement learning. This approach has the ability to learn relationships between control action such as cycle time and their effect on the vehicle queuing while pursuing a goal of maximizing intersection throughput. When this model is applied to large scale problem such as multiple phasing, will generate large state-space which limits application of conventional reinforcement learning techniques. To address this, actor- critic reinforcement learning with function approximation technique is used. The performance of the model is tested on a typical four phase four intersections arterial with variable flow is simulated using a traffic micro-simulator (VISSIM) and interfaced with the proposed model using dynamic link library (DLL). The model performance is compared with the traditional fully-actuated system. The results using this approach shows significant improvement over traditional control in the coordinated direction, especially for the traffic with demand transition.