The Distance variable was divided in two terms FAR and NEAR (Figure 2). The type of membership function used is a sigmoid with parameters slope = 1, inflection = 2.5 to the FAR set and slope = —1.8, inflection = 2.5 to the NEAR. The Distance is measured in meters. Figure 2: Near and far membership functions accordingly with the distance in meters. The Angle variable was divided in BIG and SMALL sets (Figure 3). The type of membership function is gaussian with parameters c = 0, s = 0.257 to the SMALL set and s = 0.357, c= 7 to the BIG set. The Angle is measured in radians. Figure 3: Small and big membership functions accordingly with the angle in radians. of pedestrians who are walking through the hallway. As shown by Figure 6. Where we present the first moments in the simulations above the last ones. At the first moments all the rooms are full, in the last, just the rooms located at the end of the all. Figure 6: Middle rooms are the last to evacuate. We also can note that the when the pedestrians change abruptly the direction, they start to bump one in another. This effect also happens in very crowded areas. This effect can be seen in the Figure 7. We drawn lines to show the approximate direction of movement in each place.
