APPLYING AN ARTIFICIAL NEURAL NETWORK MODEL TO FREEWAY RAMP METERING CONTROL

This paper proposes an automatic ramp metering control system that detects the current traffic situation and responds with the most suitable control strategy. An artificial neural network model is developed to compute the appropriate metering rates, taking into account several variables and their interactions in time and space dimensions. A self-retraining capability is included in the system to facilitate on-line evaluation and adjustment.