Knowledge Integration by Using Adaptive Neural- fuzzy Networks for Ramp Metering

An Adaptive Neuro-Fuzzy Inference System (ANFIS) is a neural network-based fuzzy inference system that includes combination of two soft-computing methods: Artificial Neural Networks (ANN) and fuzzy logic [1] Fuzzy logic has the ability to transform the qualitative aspects of human knowledge and insights into the process of precise quantitative analysis [2]. However, it is very problematic to transform the human thought into a rule based Fuzzy Inference System (FIS), and adequately adjust Membership Functions (MFs) of the mentioned FIS. ANFIS uses ANN’s ability of self-adaptation to the environment through the machine learning process in order to automatically adjust the MFs, and reduce the rate of errors in the determination of rules in FIS [2]. Fuzzy logic based approaches such as FIS are often used for ramp metering. Ramp metering as one of the control methods for urban motorways is formulated as the regulation of the on-ramp flow access rate into the motorway mainstream according to the several inputs. Most ramp metering algorithms based on fuzzy logic require a robust and comprehensive approach for adjusting of the FIS rule base and MFs in a complex non-linear environments such as the urban motorway traffic system.