A floor field cellular automaton for crowd evacuation considering different walking abilities

Effect of speed matching on fundamental diagram of pedestrian flow

“…Fluid dynamic and gas kinetic models [28][29][30][31] and social force models [6,[32][33][34] are typical continuous models, whereas cellular automaton (CA) models [35][36][37][38] and lattice-gas models [39][40][41][42][43] are typical discrete models. As a kind of discrete microscopic PED models, the floor field CA model, an extension of asymmetric simple exclusion process (ASEP), has its specific advantages in the extendibility, reproducing crowd self-organized phenomena, embodying individual properties, and reducing the computing complexity by translating the long-ranged interaction to local interaction [44][45][46]. It plays an important role in the study of PED.…”

Calibrating floor field cellular automaton models for pedestrian dynamics by using likelihood function optimization

“…However, the main limitation of these models is that they may not be calibrated with experimental 44 data and whether they are, they are not completely calibrated by using observed data (i.e. most of the time only few 45 parameters are calibrated). The present work introduces a methodology to estimate all the parameters included in these 46 models by defining a likelihood function and using observed trajectories (i.e.…”

“…Further assumptions can be added to the ones listened above to make the model closer to reality (see, for instance, the non-14 homogeneous cellular automata framework proposed by Was and Lubaś [43]), the different walking abilities investigated Q4 15 by Fu et al the modified dynamic floor field model proposed by Guo et al [44], Other two key issues of this modelling 16 framework that need to be discussed are conflict resolution (when more than one pedestrian compete for the same cell) and 17 multi-velocities [45]. Different solutions have been proposed to solve both issues (see Refs.…”

“…Different solutions have been proposed to solve both issues (see Refs. [27,45,46]) and can be added to 18 the general formulation described above. It is worth highlighting that the proposed methodology is applicable for different 19 types of formulation as far as the multinomial logistic assumption showed in Eq.…”

“…FERET database [54] pedestrians were engaged in their daily activities without being aware to be recorded) collected at the Informatics Forum of 43 the University of Edinburgh [60]. It is worth highlighting that any datasets could be used with the proposed methodology as 44 far as those include information about both trajectories and velocities and the time resolution is small enough to investigate 45 which cell of the pedestrian's neighbourhood is selected at each time step (see Section 3.2). 46 In the exemplary implementation, different hypothetical static floor fields have been tested (see Eq.…”