Grey Difference Model to Forecast Air Pollution in Road Tunnel

Abstract

Longitudinal ventilation system of the long tunnel in the highway is a random, sluggish and nonlinear system. To accurately predict air pollution concentration in the road tunnel is very useful and necessary for us to improve the efficiency and the quality of ventilation control system. In this paper, based on having thoroughly analyzed the physical process of the longitudinal ventilation, we have proposed a mathematic model of which the longitudinal ventilation can be described by the grey system with a grey cause and white result. By means of the grey theory, a grey prediction method to establish the discrete grey model DGM (1, 1) has been proposed to forecast the air pollutions in road tunnels. Combining with moving average smooth method, the proposed method is used to predict CO concentrations in China's Qinling No.1 tunnel separately for one minute and ten minutes. The application results show that the maximum relative error of the grey prediction method is less than 5% in one minute forecast and is less than 10% in ten minutes forecast, and the mean absolute percentage errors is only 0.89% for one minute prediction and 3.16% for ten minutes prediction.