Genetic-fuzzy logic ramp metering control for ramp metering of motorway

Abstract

This paper proposes a local responsive ramp metering algorithm based on genetic fuzzy logic control (GFLC). In this algorithm, the traffic conditions for an isolated on-ramp are responded by a knowledge-based system containing a set of membership functions and a set of fuzzy rules. In order to stabilise the traffic density at the critical density where traffic volumes reach the maximum throughput, genetic algorithm is employed to tune the parameterised membership functions. The performance of GFLC is measured by total travel time (TTT) in a simulation scenario constructed by a stochastic and microscopic simulator, Aimsun. The proposed algorithm is also compared to FLC ramp metering and no-ramp-control case to show the improvement in terms of the percentage changes of TTT. The simulation results have shown that the GFLC ramp metering provides significant improvement of TTT and better ability to maintain system flow density than the FLC ramp metering.