A note on Artificial Intelligence techniques and Dissipativity-based Approach in Traffic Signal Control For an Over-Saturated Intersection

摘要

The main motivation of this study is to take full advantage of the capabilities of neural networks techniques and dissipative system theory to design a powerful state feedback real-time controller for an over-saturated intersection. The developed controller makes a real-time decisions about whether to increase or decrease (and how much) the current green time in order to get out the over-saturation situation. First, a discrete-time model that describes the evolution of the queue lengths at signalized intersection is presented. Then, the control problem is formulated and solved by using the dissipative system theory. Moreover, the proposed intersection controller needs to be provided with a real-time input traffic flow data. To achieve this goal, we use the artificial neural networks technique as input traffic flow data predictor. The results of the simulations indicate that our control strategy guarantees a high degree of control benefit.