Facility Design for Pedestrian Evacuation Using Cellular Automation

A cellular automation (CA) model of pedestrian movements with 3D visualization is implemented for facility planning. The CA methodology is efficient for large-scale and repeated simulations and describes pedestrian behaviors adequately. Different from the previous studies that attempt to calibrate their models at microscopic level or ignore model calibration completely, this paper calibrates the model with data at macroscopic level (flow, density, and speed). The calibrated pedestrian model is useful for optimal facility design because it reflects the complicated interactions between pedestrians and the facility. After candidate facility designs are evaluated with the pedestrian models, a two-stage procedure is adopted for the statistical selection of the alternatives. The numerical example of a transportation terminal shows that the best design can be selected with a pre-determined probability of correct selection and targeted difference between the best and the second best systems. Moreover, required numbers of repetitions are calculated and conducted to avoid incorrect selection or unnecessary repetitions. Therefore, the procedure proposed is an effective and efficient tool for facility design.