Abstract: Memory affects driving behaviors of drivers such as visual search and route planning,and then influences efficiency and safety of road traffic.In order to describe characteristics of drivers′memory variation under repeated driv-ing situations,a driving simulation experiment is designed.Cumulative stimulus of repeated driving in a same scene is studied.A memory scale is adopted to measure memory degrees,and dynamic relationship between memory growth of drivers and the number of repeated driving is analyzed.Models of memory growth of drivers under cumulative stimulus are developed by using Mitscherlich,Modified Weibull,and Richards function,respectively.In addition,fitting effects of these models are compared by overall evaluation indices of adjusted determination coefficient,sum of the squared errors, and root mean square error.The results show that these three models can effectively describe the characteristics of driv-ers′memory growth.In conclusion,the model using Richards function has the best precision,of which the average adjus-ted R-square is 0.988 4,which reveals the essence of memory assimilation and dissimilation.It fits for being applied to study memory growth of drivers under repeated driving situations.