Design of a genetic fuzzy approach for ramp metering

This paper proposes a nonlinear approach for designing traffic-responsive ramp controls using a genetic fuzzy approach. The problem is formulated as a nonlinear feedback control problem. To overcome the conventional problems of the calibration process of fuzzy controllers and improve the overall performance of ramp metering, an adaptive genetic-based algorithm is integrated into the system to periodically tune the fuzzy sets parameters. The approach thus adapts the control system automatically to changing traffic patterns. The objective of the ramp control is to minimize the total time spent in the freeway system while maintaining acceptable ramp service levels. Traffic data from a ramp study site in the Munich Autobahn (A9 motorway) was used to assess the genetic fuzzy controller using a hydrodynamic traffic model to estimate the genetic fitness. The paper concludes that adaptive fuzzy control based on genetic algorithms is expected to enhance the performance of ramp metering without compromising the cost-effectiveness associated with fuzzy controllers