Fuzzy Neural Network Model Applied in the Traffic Flow Prediction

The paper proposes a fuzzy neural network model (FNNM) strategy for predicting the traffic flow of real time traffic control systems. The proposed model is composed of two modular. One is a fuzzy network (FN), which is used for fuzzy clustering. Each cluster represents one kind of specific traffic pattern. The other is a neural network (NN), which is one-layer network and is used for partitioning the relationship of input and output vector. And the FN module supervises the learning of the NN. That is, the features of the traffic samples are employed to guide the training of the NN. Moreover, an online iterative predictive algorithm is presented in this paper to predict the traffic flow according to the sampled data of the upstream cross roads. Finally, the real sampled traffic flow data is employed to validate the proposed method. Results show that the proposed traffic flow prediction strategy based on fuzzy neural network model is feasible and effective