Pollution level estimation for tunnel ventilation control

Traffic tunnels built in recent years are equipped with traffic counters and pollution sensors. Utilizing these built-in sensors, it is possible to develop an algorithm to estimate the amount of pollutants exhausted from the various types of vehicles passing through the tunnel. Also, with this estimated data and the sensor outputs, more accurate pollution levels can be assessed utilizing a Kalman filter. The diffusion of pollutants in a tunnel can be described with a one-dimensional diffusion and advection equation. This equation is approximated with interpolation functions and a weighted residual method converts to an adequate form for. standard state estimate algorithms. With this converted equation, a least square based, algorithm is developed, whose outputs are the estimated amounts of pollutants emitted from each type of vehicles. Also, a Kalman filter is utilized to more accurately estimate the pollution levels of the tunnel under the existence of model and measurement uncertainties. In order, to verify the feasibility of the developed algorithms, experiments and simulations are performed. The real data is acquisitioned from the Dunnae tunnel located in Young-Dong highway in Korea. The estimated emission rates and the computer simulated concentration levels agree reasonably to the measured values.