| As the economy develops in China,more and more large-scale recreational activities in public places are being hold which attract lots of people.Nowadays crowd gatherings are very common phenomena.How to ensure the safety of people in public places has become one of the hot spots.In the past,most of traditional research use simulation or build kinds of mathematical models to study the characteristics from macro to micro perspectives.And some analyze the causes of some specific accident and summarized relevant management recommendations.However,these studies belong to the type of verification research.They are all carried on after accidents and test the correction of their studies.This dissertation breaks through traditional methods.Experiments are conducted on the metro in Shanghai during the morning peak time to collect the physical parameters such as blood oxygen saturation,pulse rate and the end-tide carbon dioxide tension to study the different pedestrians’ states in extreme crowding environments.But in the end,only pulse rate is used for data analysis.By comparing and analyzing the features of data,we provide a new idea for improving the safety management of the public places.The wavelet denoising method is adapted to eliminate the inadequate data.And a new threshold function is proposed to improve the efficiency of the algorithm making up for the shortcomings of the traditional soft threshold and hard threshold.Linear spectral analysis,one-way analysis of variance,wavelet packet scaling and some linear parameters are used to quantitatively and qualitatively showing the differences among these data.Besides,in order to verify the chaotic characteristics of physiological signals,the correlation dimension,maximum lyapunov exponent,approximate entropy and complexity of physiological signals under different states are selected and calculated.The results show the complexity and non-linearity of the pulse signals.Experiment results show that the states of the human body can be effectively reflected by the physiological signals.When the external environmental stimulates the human body,different individuals have different responses due to different sex,age and physical health.But healthy individuals have some adaptability towards to the outside environment.Although the signals fluctuate frequently,the physiological state of an individual will not be affected.Still,compared with women and men,physiological signals change more intensely.However,the results also demonstrate when an individual is in an uncomfortable condition,the signals change abnormally and it is more sensitive to the external environment stimuli.This conclusion directs a new way for public safety management,that is,in the public areas,the identifications of high-vulnerability groups are very crucial for making sure of the safety of the crowd. |