Short-time traffic flow prediction based on seasonal gray Fourier model

To handle the periodic, nonlinear, and stochastic aspects of short-time traffic flow data, a seasonal gray Fourier model based on the complex Simpson formula is proposed. The seasonal GM (1, 1) model is used to optimize the background values first, and then the prediction results are adjusted using the Fourier series method. The new model was applied to the prediction of traffic flow on Whitemud Drive in Canada, and the numerical results indicated that the new model's mean absolute percentage error was 1.54 percent and its fit was 0.996, which were significantly better than those of the traditional GM (1, 1) model, the seasonal GM (1, 1) model, and the Fourier optimized seasonal GM (1, 1) model.