Design of a real-time controller for an over-saturated intersection using self-organizing maps and dissipativity concept

The dissipativity, which is related to the notion of energy, is a very important concept in systems and control theory. It is commonly used in physical sciences, but with limited application in transportation systems. In this paper, we propose a real-time state feedback control strategy based on the dissipativity concept for an over-saturated intersection. The developed controller adjusts the green time of each cycle in order to get out of the over-saturation situation. For the proposed controller to be effective, it must have an accurate prediction of the input traffic flow over a short periods. To achieve this goal, the Kohonen Self-Organizing Map (SOM) neural network is used to predict the one cycle ahead input traffic flow. The simulation results show the effectiveness of the proposed control strategy.