Neural Network Optimized by Genetic Algorithm of Models for Real-Time Forecast of Traffic Flow

Abstract

Short-term traffic flow forecast plays an important role in transit scheduling. A high-order generalized neural network model is constructed to actualize dynamic forecast on-line and a hybrid genetic algorithm and identical dimension recurrence idea are performed to optimize the structure and shape of neural network dynamically so as to enhance its forecast accuracy. With data collected from Dazhi Str., Harbin as the system input, the experimental result indicates that the average relative error of forecast is 5.53% and the maximum is less than 21%, which proves that the proposed neural network model can satisfy the precision request, accelerate the convergence speed, improve the global generalization ability and possess the practicality in short-term traffic flow forecast.