Study On Pedestrians’ Spatial-temporal Data Collection Method And Walking Behavior

Study on pedestrian walking behavior is an important topic in emergency managementand life protection. Due to randomness of walker’s movement, difference of variouspedestrians and impact of environment, study on practical models of crowds movingbehavior is still in its infancy. One of restrictions of pedestrian movement modeling is thedata collection for pedestrian dynamic. Traditional data collection approaches, such asvideo detection and wireless location, due to its poor anti-interference performance or highcost has not been expanded in pedestrian walking models estimation and validation. Tocollect each walker’s moving characteristic data, we apply techniques of radio frequencyidentification and opportunistic network to scale the spatial coordinates and use the onlytimer in field to maintain time synchronization. By detecting each node’s position, i.e.,pedestrian individual’s walking trajectory, we obtain each walker’s moving characteristicdata in crowed state. Also, we develop a specialized data analytic program to merge anddecode original data. We then verify the whole pedestrian movement data collection systemby simulating channel experimental scenario. It is found that this proposal is not onlycreative but reliable for pedestrian’s walking behavior data collection. It is hoped that theidea of this study can be helpful in validating the theoretical models proposed up to nowand promoting the crowd emergency management.Based on the new approach of individual pedestrian’s walking temporal-spatialparameters data collection system, we collect the individual pedestrian’s walking behaviordata by simulating pedestrians’ walking scenario through channels. Then we analyzepedestrian’ walking behavior on macro and micro level respectively. Applying the Levenehomogeneity test of variance, significant analysis of pedestrians’ walking behaviorimpacted by the scenario of channel and the state of crowd are tested. And the volatility ofthe trajectory of pedestrians’ movement through channels is fitted through the pedestrianwalking offset statistical distribution. It is shown that the influence on pedestrians’ walkingbehavior by disturbing terms (high density, obstacles and bottlenecks) under the channelscenario is markedly. We also put forward a new pedestrian movement model based on theresearches of field’s paradigm in physics and pedestrians’ social force model, and we call itpedestrian moving field model. The model can be described as basic driving field, humanfield, obstacle field and attraction field’s superposition effect. Each typical sub-field’seffection on pedestrian’s walking behavior is proposed through pedestrian moving field strength and potential distribution. Our new pedestrian moving field model is successfullyis used to validated and estimated using typical channel experimental scenario collectiondata. The pedestrian moving field model can do well to describe and study the pedestrians’walking behavior from the perspective of potential distribution.