Abstract: Cyclist safety has presently become an increasingly severe issue. We address cyclist riding stability with cycling data and traffic environment data collected in cyclist riding experiment. We also investigate the influential factors affecting riding stability based on cycling data and traffic environment data with corner changing rate as evaluation index for riding stability.We analyze data with Bayesian inference based random parameter model to consider the heterogeneity of individuals. Analytical results indicate that cyclist riding stability is only affected by riding speed variance coefficient.The greater the coefficient is, the greater corner changing rate is.This shows worse riding stability. Individual difference considered random parameter model has better degree of fitting, as compared with conventional fixed parameter models.

Key words: cyclist safety, riding experiment, Bayesian inference, random parameter model