Modeling and simulation of pedestrian dynamical behavior based on a fuzzy logic approach

This study proposes a fuzzy logic approach to model and simulate pedestrian dynamical behaviors, which takes full advantage of human experience and knowledge and perceptual information obtained from interactions with surrounding environments. First, the radial-based method is adopted to represent the physical space. A pedestrian’s visual field, defined as a fan-shaped area with a certain visual distance and visual angle, is divided into five sectors. Then, the motion states of a pedestrian are determined by the integration of recommendations of local obstacle-avoiding behavior, regional path-searching behavior and global goal-seeking behavior with mutable weighting factors at three different scopes. These elementary behaviors and weighting’s assignment principle are modeled as fuzzy inference systems with the input information of a pedestrian’s perception toward surrounding environments. A pedestrian is guided to avoid the front obstacles and select the lowest negative energy path by local obstacle-avoiding behavior and regional path-searching behavior, respectively. The global goal-seeking behavior makes a pedestrian has a tendency of moving in direction of his/her goal regardless of external environments. The magnitudes of weighting factors are adjusted automatically to coordinate three elementary behaviors and resolve potential conflicts. At last, the effectiveness of the proposed model is validated by simulations of crowd evacuation, unidirectional and bidirectional pedestrian flows. The simulation results are analyzed from both qualitative and quantitative aspects, which indicate that the fuzzy logic based pedestrian model can get true reappearance of self-organization phenomena such as ‘arching and clogging’, ‘faster-is-slower effect’ and ‘lane formation’, and the fundamental diagrams are in matching with a large variety of empirical and experimental data. A further study finds that walking habits have negligible influence on the fundamental diagrams of bidirectional pedestrian flow at least for densities of ρ < 3p/m2.