A Hybrid Grey Markov Prediction Model for Traffic Volume

A hybrid Grey Markov prediction model was developed by incorporating Markov forecasting model into grey forecasting model, and it was used to address the problem of traffic volume prediction. This compound model predicts me general trend of original data with unified-dimension new message GM(1,1), then applies first order Markov model to forecast the relative error series, finally use the forecasting relative error data to amend GM(1,1) prediction results. To confirm the applicability of this model, we employed Grey Markov forecasting model to forecast real time traffic volume of an intersection in SUZHOU city. As verified in experimental results, the Grey Markov model is overmatched GM(1,1) for traffic volume forecasting.