AN OPEN ACCESS JOURNAL
 JJCE Submission Home


Neuro-fuzzy -based Crowd Speed Analysis At Mass Gathering Events

Submitted2019-08-27
Last Update2019-08-27
TitleNeuro-fuzzy -based Crowd Speed Analysis At Mass Gathering Events
Author(s)Author #1
Author title:
Name: Poojari Yugendar
Org: Research Scholar, Transportation Division, Department of Civil Engineering, National Institute of Technology Warangal-506004, Telangana State, India.
Country:
Email: poojariyugendar1@gmail.com

Author #2
Author title:
Name: K.V.R. Ravishankar
Org: Assistant Professor, Transportation Division, Department of Civil Engineering, National Institute of Technology Warangal-506004, Telangana State, India
Country:
Email: kvrrshankar@gmail.com

Other Author(s)
Contact AuthorAuthor #1
Alt Email: poojariyugendar1@gmail.com
Telephone:
KeywordsCrowd, ANN, ANFIS, MLR, RMSE, MAPE, Speed.
AbstractAirports, shopping centers, sport stadiums and religious houses,� etc. are largely crowded areas. There is a need for the design and planning of crowded facilities to handle large volumes of crowd. Injuries and fatalities in emergency evacuations were not only caused by the hazards, but also by actions of the crowd. Stampedes are caused both by the real hazards like fire, earthquake,� etc. and the behavior of the crowd. Crowd speed is one major factor in analyzing crowd events. The physical factors and environmental factors influence the speed of an individual in a crowd. In this study, effects of factors like age, gender, group size, child holding, child carrying, people with luggage and without luggage on crowd speed are considered for analysis. The statistical analysis concluded that there was a significant effect of age, gender, density and luggage on the crowd walking speed. Multi-linear regression (MLR), Artificial Neural Networks (ANNs) and Adaptive Neuro-fuzzy Inference System (ANFIS) models were developed between crowd speed and significant factors observed from the statistical analysis. The results showed that the ANFIS model results are the best fitted compared to other models. The Mean Absolute Percentage Error (MAPE) and Route Mean Square Error (RMSE) of the ANFIS model are determined as 0.130 and 0.098.
Paperview paper 4772.pdf (624KB)

http://www.just.edu.jo